conductor(video_analysis): Initialize 12 child + 1 synthesis spec scaffolds
Each child spec is lightweight (~100 lines): references the umbrella, gives video details, specifies the 7 deliverables (transcript.json, frames/, ocr.md, report.md 1000-10000 LOC, summary.md), and the 5-phase pipeline. Children in execution order: 1. cs229_building_llms (Stanford CS229, Cluster E) 2. probability_logic (Cluster A) 3. entropy_epiplexity (Cluster A) 4. score_dynamics_giorgini (Cluster A) 5. platonic_intelligence_kumar (Cluster B) 6. free_lunches_levin (Cluster B) 7. generic_systems_fields (Cluster C) 8. brain_counterintuitive (Cluster C) 9. neural_dynamics_miller (Cluster C) 10. multiscale_hoffman (Cluster C) 11. cs336_architectures (Stanford CS336, Cluster E) 12. creikey_dl_cv (Cluster D) Plus 1 synthesis track (video_analysis_synthesis_20260621) blocked_by all 12 children.
This commit is contained in:
@@ -0,0 +1,91 @@
|
||||
# Track: Video Analysis — Most Counterintuitive Way to Build a Brain
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** C (Biological / cognitive / generic systems)
|
||||
|
||||
> **Parent:** Child #8 of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | The Most Counterintuitive Way to Build a Brain |
|
||||
| **Author** | (unknown — verify during execution) |
|
||||
| **URL** | https://youtu.be/cDxtFtoQVNc |
|
||||
| **Cluster** | C |
|
||||
| **Slug** | `brain_counterintuitive` |
|
||||
| **Execution order** | #8 of 12 (concrete biological, after #7 meta-frame) |
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Frames | `artifacts/frames/*.jpg` | 50-500 |
|
||||
| Extraction meta | `artifacts/extraction_meta.json` | Frame paths + hashes |
|
||||
| OCR | `artifacts/ocr.md` | Full OCR per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC** |
|
||||
| Summary | `summary.md` | 200-400 words |
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline
|
||||
|
||||
- [ ] **Phase 1:** Acquire
|
||||
- [ ] **Phase 2:** Keyframes
|
||||
- [ ] **Phase 3:** OCR
|
||||
- [ ] **Phase 4:** Synthesis (1000-10000 LOC)
|
||||
- [ ] **Phase 5:** Verification
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure
|
||||
|
||||
8 sections per umbrella spec §FR6.
|
||||
|
||||
```
|
||||
# The Most Counterintuitive Way to Build a Brain
|
||||
**Source:** https://youtu.be/cDxtFtoQVNc
|
||||
**Author:** <verify>
|
||||
**Cluster:** C
|
||||
**Slug:** brain_counterintuitive
|
||||
|
||||
## 1. TL;DR
|
||||
## 2. Key Concepts ← expect unconventional neuroscience, biological computation
|
||||
## 3. Frame Analysis
|
||||
## 4. Transcript Highlights
|
||||
## 5. Mathematical / Theoretical Content
|
||||
## 6. Connections
|
||||
## 7. Open Questions
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections
|
||||
|
||||
- **Forward to:** `neural_dynamics_miller` (more conventional neuro), `multiscale_hoffman` (synthesis).
|
||||
- **Backward from:** `generic_systems_fields` (meta-frame), `free_lunches_levin` (agential materials).
|
||||
- **Likely rich cross-references:** `neural_dynamics_miller` (most direct — both about brain/cognition).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification
|
||||
|
||||
- [ ] All 7 deliverables present
|
||||
- [ ] `report.md` 1000-10000 LOC
|
||||
- [ ] Tests pass
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md)
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md)
|
||||
@@ -0,0 +1,91 @@
|
||||
# Track: Video Analysis — Creikey DL/CV for Game Developers
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** D (Applied / practical)
|
||||
|
||||
> **Parent:** Child #12 (last) of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | Creikey — Deep Learning and Computer Vision for Game Developers — BSC 2025 |
|
||||
| **Author** | Creikey |
|
||||
| **URL** | https://youtu.be/yxkUvXs-hoQ |
|
||||
| **Cluster** | D (Applied / practical) |
|
||||
| **Slug** | `creikey_dl_cv` |
|
||||
| **Execution order** | #12 of 12 (applied capstone — validates the theory against practice) |
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Frames | `artifacts/frames/*.jpg` | 50-500 |
|
||||
| Extraction meta | `artifacts/extraction_meta.json` | Frame paths + hashes |
|
||||
| OCR | `artifacts/ocr.md` | Full OCR per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC markdown** |
|
||||
| Summary | `summary.md` | 200-400 words |
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline (5 phases)
|
||||
|
||||
- [ ] **Phase 1:** Acquire
|
||||
- [ ] **Phase 2:** Keyframes
|
||||
- [ ] **Phase 3:** OCR
|
||||
- [ ] **Phase 4:** Synthesis (1000-10000 LOC)
|
||||
- [ ] **Phase 5:** Verification
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure
|
||||
|
||||
8 sections per umbrella spec §FR6.
|
||||
|
||||
```
|
||||
# Creikey — Deep Learning and Computer Vision for Game Developers (BSC 2025)
|
||||
**Source:** https://youtu.be/yxkUvXs-hoQ
|
||||
**Author:** Creikey
|
||||
**Cluster:** D
|
||||
**Slug:** creikey_dl_cv
|
||||
|
||||
## 1. TL;DR
|
||||
## 2. Key Concepts ← expect game-specific DL/CV applications, real-time constraints
|
||||
## 3. Frame Analysis
|
||||
## 4. Transcript Highlights
|
||||
## 5. Mathematical / Theoretical Content
|
||||
## 6. Connections ← should connect back to CS229 + CS336 + math foundations
|
||||
## 7. Open Questions
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections
|
||||
|
||||
- **Forward to:** none (last video in campaign; the synthesis track comes after).
|
||||
- **Backward from:** all 11 prior videos — this is the capstone that validates the theory against practice.
|
||||
- **Likely rich cross-references:** `cs336_architectures` (architectures used in practice), `cs229_building_llms` (foundational ML), `score_dynamics_giorgini` (training applied).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification
|
||||
|
||||
- [ ] All 7 deliverables present
|
||||
- [ ] `report.md` 1000-10000 LOC
|
||||
- [ ] Tests pass
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md)
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md)
|
||||
@@ -0,0 +1,105 @@
|
||||
# Track: Video Analysis — Stanford CS229 (Building LLMs)
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** E (Stanford course VODs >1hr)
|
||||
|
||||
> **Parent:** This is child #1 of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella. See [umbrella spec.md](../../video_analysis_campaign_20260621/spec.md) for the full design, multi-pass context (Pass 2 = de-obfuscation, Pass 3 = projection), and tooling requirements.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | Stanford CS229 — Machine Learning — Building Large Language Models (LLMs) |
|
||||
| **Author** | Stanford CS229 |
|
||||
| **URL** | https://youtu.be/9vM4p9NN0Ts |
|
||||
| **Cluster** | E (Stanford course VODs >1hr) |
|
||||
| **Estimated duration** | >1hr (Stanford course lecture) |
|
||||
| **Slug** | `cs229_building_llms` |
|
||||
| **Execution order** | #1 of 12 (canonical ML foundation — sets vocabulary for everything after) |
|
||||
|
||||
**Pre-execution note (2026-06-21):** This video's oEmbed API fetch returned 401. This may indicate a private/age-restricted video; `yt-dlp` may still work. **Phase 1 of this track must verify yt-dlp access before downloading the mp4.** If `yt-dlp` also fails, fall back to manual transcript sourcing (if available) or escalate.
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript (timestamped + plain) | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Extracted unique frames | `artifacts/frames/*.jpg` | 50-500 frames |
|
||||
| Extraction metadata | `artifacts/extraction_meta.json` | Frame paths + hashes + timestamps |
|
||||
| OCR results | `artifacts/ocr.md` | Full OCR text per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC markdown** |
|
||||
| Quick summary | `summary.md` | 200-400 words |
|
||||
|
||||
**Optional (added per child track execution convention):** `plan.md`, `metadata.json`, `state.toml`.
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline (5 phases)
|
||||
|
||||
Per the umbrella spec §FR5. Each phase commits atomically.
|
||||
|
||||
- [ ] **Phase 1: Acquire.** Run `extract_transcript.py` (always succeeds, fast). Run `download_video.py` if keyframe extraction needs the video. Verify `yt-dlp` access first.
|
||||
- [ ] **Phase 2: Keyframes.** Run `extract_keyframes.py` with threshold 0.4. Manually review frame set; flag candidates that look wrong.
|
||||
- [ ] **Phase 3: OCR.** Run `ocr_frames.py` on `frames/`. Spot-check OCR quality.
|
||||
- [ ] **Phase 4: Synthesis.** Tier 3 worker prompt: transcript + OCR + frame images → `report.md`. Human review + iteration. Target 1000-10000 LOC.
|
||||
- [ ] **Phase 5: Verification.** Idempotency check. Audit checklist. End-of-track report.
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure (8 sections)
|
||||
|
||||
Per umbrella spec §FR6. Target: 1000-10000 LOC.
|
||||
|
||||
```
|
||||
# Stanford CS229 — Building Large Language Models (LLMs)
|
||||
**Source:** https://youtu.be/9vM4p9NN0Ts
|
||||
**Author:** Stanford CS229
|
||||
**Date Added to Campaign:** 2026-06-21
|
||||
**Cluster:** E
|
||||
**Slug:** cs229_building_llms
|
||||
|
||||
## 1. TL;DR (3-5 sentences)
|
||||
## 2. Key Concepts (5-15 bullets, each with brief explanation)
|
||||
## 3. Frame Analysis (one subsection per significant frame; embed image; describe visual + OCR + significance)
|
||||
## 4. Transcript Highlights (with timestamps; verbatim quotes of key passages)
|
||||
## 5. Mathematical / Theoretical Content (formal notation; derivations; references)
|
||||
## 6. Connections to Other Videos in Campaign (forward + backward links)
|
||||
## 7. Open Questions / Follow-up
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections (forward + backward)
|
||||
|
||||
- **Forward to:** everything else in the campaign — this is the canonical ML/LLM foundation.
|
||||
- **Backward from:** none (this is video #1).
|
||||
- **Likely rich cross-references:** `cs336_architectures` (later in the campaign, deep dive on transformer architectures; this video provides the why).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification (per umbrella spec §12)
|
||||
|
||||
- [ ] All 7 deliverable artifacts present
|
||||
- [ ] `report.md` is 1000-10000 LOC
|
||||
- [ ] `summary.md` is 200-400 words
|
||||
- [ ] All 8 report sections populated
|
||||
- [ ] Idempotency check passes
|
||||
- [ ] Tests pass
|
||||
- [ ] Per-task commits with git notes
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md) — full design
|
||||
- [Umbrella plan.md](../../video_analysis_campaign_20260621/plan.md) — campaign-level plan
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md) — child index
|
||||
- [Umbrella metadata.json](../../video_analysis_campaign_20260621/metadata.json) — scope + risk register
|
||||
@@ -0,0 +1,96 @@
|
||||
# Track: Video Analysis — Stanford CS336 Lecture 3: Architectures
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** E (Stanford course VODs >1hr)
|
||||
|
||||
> **Parent:** Child #11 of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella. See [umbrella spec.md](../../video_analysis_campaign_20260621/spec.md) for full design.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | Stanford CS336 — Language Modeling from Scratch, Spring 2026, Lecture 3: Architectures |
|
||||
| **Author** | Stanford CS336 Spring 2026 |
|
||||
| **URL** | https://youtu.be/lVynu4bo1rY |
|
||||
| **Cluster** | E (Stanford course VODs >1hr) |
|
||||
| **Estimated duration** | >1hr (Stanford course lecture) |
|
||||
| **Slug** | `cs336_architectures` |
|
||||
| **Execution order** | #11 of 12 (deep dive on transformer architectures; pairs back to CS229 with full context from prior videos) |
|
||||
|
||||
**Pre-execution note (2026-06-21):** This video's oEmbed API fetch returned 401. This may indicate a private/age-restricted video; `yt-dlp` may still work. **Phase 1 of this track must verify yt-dlp access before downloading the mp4.** If `yt-dlp` also fails, fall back to manual transcript sourcing (if available) or escalate.
|
||||
|
||||
**Position rationale:** Placed late (after Clusters A/B/C) rather than early so the Tier 3 worker has full context from CS229 + math + Platonic + biological videos when analyzing the architecture details. CS229 (#1) sets the "why" for transformer architectures; CS336 (#11) is the "how" deep dive.
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Frames | `artifacts/frames/*.jpg` | 50-500 |
|
||||
| Extraction meta | `artifacts/extraction_meta.json` | Frame paths + hashes |
|
||||
| OCR | `artifacts/ocr.md` | Full OCR per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC markdown** |
|
||||
| Summary | `summary.md` | 200-400 words |
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline (5 phases)
|
||||
|
||||
- [ ] **Phase 1: Acquire.** Run `extract_transcript.py` + `download_video.py`. Verify `yt-dlp` access first.
|
||||
- [ ] **Phase 2: Keyframes.** `extract_keyframes.py` with threshold 0.4.
|
||||
- [ ] **Phase 3: OCR.** `ocr_frames.py`.
|
||||
- [ ] **Phase 4: Synthesis.** Tier 3 worker: transcript + OCR + frames → `report.md` (1000-10000 LOC).
|
||||
- [ ] **Phase 5: Verification.** Idempotency + audit + end-of-track report.
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure
|
||||
|
||||
8 sections per umbrella spec §FR6.
|
||||
|
||||
```
|
||||
# Stanford CS336 Lecture 3: Architectures
|
||||
**Source:** https://youtu.be/lVynu4bo1rY
|
||||
**Author:** Stanford CS336 Spring 2026
|
||||
**Cluster:** E
|
||||
**Slug:** cs336_architectures
|
||||
|
||||
## 1. TL;DR
|
||||
## 2. Key Concepts ← expect transformer architectures, attention, FFN, MoE, etc.
|
||||
## 3. Frame Analysis ← expect dense formulas/diagrams; OCR critical
|
||||
## 4. Transcript Highlights
|
||||
## 5. Mathematical / Theoretical Content ← dominant
|
||||
## 6. Connections ← heavy cross-refs to CS229 + prior cluster videos
|
||||
## 7. Open Questions
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections
|
||||
|
||||
- **Forward to:** `creikey_dl_cv` (applied capstone — uses these architectures in games).
|
||||
- **Backward from:** `cs229_building_llms` (sets "why transformer architectures"), `score_dynamics_giorgini` (training dynamics), `platonic_intelligence_kumar` (representations inside the architecture).
|
||||
- **Likely rich cross-references:** `cs229_building_llms` (most direct — same LLM topic, different depth), `platonic_intelligence_kumar` (representations inside architectures).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification
|
||||
|
||||
- [ ] All 7 deliverables present
|
||||
- [ ] `report.md` 1000-10000 LOC
|
||||
- [ ] Tests pass
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md)
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md)
|
||||
@@ -0,0 +1,91 @@
|
||||
# Track: Video Analysis — From Entropy to Epiplexity (Wilson & Finzi)
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** A (Math & information-theoretic foundations)
|
||||
|
||||
> **Parent:** Child #3 of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | From Entropy to Epiplexity |
|
||||
| **Author** | Andrew Wilson and Marc Finzi |
|
||||
| **URL** | https://youtu.be/_U8AwUq_aJQ |
|
||||
| **Cluster** | A |
|
||||
| **Slug** | `entropy_epiplexity` |
|
||||
| **Execution order** | #3 of 12 (builds on entropy; pairs with #2) |
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Frames | `artifacts/frames/*.jpg` | 50-500 |
|
||||
| Extraction meta | `artifacts/extraction_meta.json` | Frame paths + hashes |
|
||||
| OCR | `artifacts/ocr.md` | Full OCR per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC** |
|
||||
| Summary | `summary.md` | 200-400 words |
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline
|
||||
|
||||
- [ ] **Phase 1:** Acquire
|
||||
- [ ] **Phase 2:** Keyframes
|
||||
- [ ] **Phase 3:** OCR
|
||||
- [ ] **Phase 4:** Synthesis (1000-10000 LOC report)
|
||||
- [ ] **Phase 5:** Verification
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure
|
||||
|
||||
8 sections per umbrella spec §FR6.
|
||||
|
||||
```
|
||||
# From Entropy to Epiplexity
|
||||
**Source:** https://youtu.be/_U8AwUq_aJQ
|
||||
**Author:** Andrew Wilson and Marc Finzi
|
||||
**Cluster:** A
|
||||
**Slug:** entropy_epiplexity
|
||||
|
||||
## 1. TL;DR
|
||||
## 2. Key Concepts ← expect epiplexity vs entropy, model complexity measures
|
||||
## 3. Frame Analysis
|
||||
## 4. Transcript Highlights
|
||||
## 5. Mathematical / Theoretical Content ← dominant
|
||||
## 6. Connections
|
||||
## 7. Open Questions
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections
|
||||
|
||||
- **Forward to:** `score_dynamics_giorgini` (learning dynamics), `cs336_architectures` (model complexity in LLMs).
|
||||
- **Backward from:** `probability_logic` (extends logic with probability).
|
||||
- **Likely rich cross-references:** `probability_logic` (most direct — both foundational).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification
|
||||
|
||||
- [ ] All 7 deliverables present
|
||||
- [ ] `report.md` 1000-10000 LOC
|
||||
- [ ] Tests pass
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md)
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md)
|
||||
@@ -0,0 +1,91 @@
|
||||
# Track: Video Analysis — Free Lunches (Levin)
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** B (Platonic / geometric AI representations)
|
||||
|
||||
> **Parent:** Child #6 of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | Free Lunches: Model Systems for Studying the Agential Gifts from the Platonic Space |
|
||||
| **Author** | Michael Levin |
|
||||
| **URL** | https://youtu.be/K8BmMU1Tm-I |
|
||||
| **Cluster** | B |
|
||||
| **Slug** | `free_lunches_levin` |
|
||||
| **Execution order** | #6 of 12 (pairs with #5; agential/Platonic synthesis) |
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Frames | `artifacts/frames/*.jpg` | 50-500 |
|
||||
| Extraction meta | `artifacts/extraction_meta.json` | Frame paths + hashes |
|
||||
| OCR | `artifacts/ocr.md` | Full OCR per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC** |
|
||||
| Summary | `summary.md` | 200-400 words |
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline
|
||||
|
||||
- [ ] **Phase 1:** Acquire
|
||||
- [ ] **Phase 2:** Keyframes
|
||||
- [ ] **Phase 3:** OCR
|
||||
- [ ] **Phase 4:** Synthesis (1000-10000 LOC)
|
||||
- [ ] **Phase 5:** Verification
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure
|
||||
|
||||
8 sections per umbrella spec §FR6.
|
||||
|
||||
```
|
||||
# Free Lunches: Model Systems for Studying the Agential Gifts from the Platonic Space
|
||||
**Source:** https://youtu.be/K8BmMU1Tm-I
|
||||
**Author:** Michael Levin
|
||||
**Cluster:** B
|
||||
**Slug:** free_lunches_levin
|
||||
|
||||
## 1. TL;DR
|
||||
## 2. Key Concepts ← expect agential materials, basal cognition, Platonic space, model systems
|
||||
## 3. Frame Analysis
|
||||
## 4. Transcript Highlights
|
||||
## 5. Mathematical / Theoretical Content ← dominant
|
||||
## 6. Connections
|
||||
## 7. Open Questions
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections
|
||||
|
||||
- **Forward to:** `generic_systems_fields` (crosses into Cluster C — generic behavior), `brain_counterintuitive` (agential materials), `multiscale_hoffman` (collective intelligence).
|
||||
- **Backward from:** `platonic_intelligence_kumar` (Platonic representations, agential lens).
|
||||
- **Likely rich cross-references:** `platonic_intelligence_kumar` (most direct — both Platonic), `multiscale_hoffman` (collective intelligence).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification
|
||||
|
||||
- [ ] All 7 deliverables present
|
||||
- [ ] `report.md` 1000-10000 LOC
|
||||
- [ ] Tests pass
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md)
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md)
|
||||
@@ -0,0 +1,91 @@
|
||||
# Track: Video Analysis — Interesting Behavior by Generic Systems (Fields)
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** C (Biological / cognitive / generic systems)
|
||||
|
||||
> **Parent:** Child #7 of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | Interesting Behavior by Generic Systems |
|
||||
| **Author** | Chris Fields |
|
||||
| **URL** | https://youtu.be/QeMajYvhEbI |
|
||||
| **Cluster** | C |
|
||||
| **Slug** | `generic_systems_fields` |
|
||||
| **Execution order** | #7 of 12 (meta-theoretical framing for Cluster C) |
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Frames | `artifacts/frames/*.jpg` | 50-500 |
|
||||
| Extraction meta | `artifacts/extraction_meta.json` | Frame paths + hashes |
|
||||
| OCR | `artifacts/ocr.md` | Full OCR per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC** |
|
||||
| Summary | `summary.md` | 200-400 words |
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline
|
||||
|
||||
- [ ] **Phase 1:** Acquire
|
||||
- [ ] **Phase 2:** Keyframes
|
||||
- [ ] **Phase 3:** OCR
|
||||
- [ ] **Phase 4:** Synthesis (1000-10000 LOC)
|
||||
- [ ] **Phase 5:** Verification
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure
|
||||
|
||||
8 sections per umbrella spec §FR6.
|
||||
|
||||
```
|
||||
# Interesting Behavior by Generic Systems
|
||||
**Source:** https://youtu.be/QeMajYvhEbI
|
||||
**Author:** Chris Fields
|
||||
**Cluster:** C
|
||||
**Slug:** generic_systems_fields
|
||||
|
||||
## 1. TL;DR
|
||||
## 2. Key Concepts ← expect generic systems, emergence, interesting behavior, information-theoretic life
|
||||
## 3. Frame Analysis
|
||||
## 4. Transcript Highlights
|
||||
## 5. Mathematical / Theoretical Content ← dominant
|
||||
## 6. Connections
|
||||
## 7. Open Questions
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections
|
||||
|
||||
- **Forward to:** `brain_counterintuitive` (concrete biological example), `neural_dynamics_miller` (concrete neuro), `multiscale_hoffman` (synthesis).
|
||||
- **Backward from:** `free_lunches_levin` (agential + Platonic, transitions into biology).
|
||||
- **Likely rich cross-references:** `multiscale_hoffman` (cross-cluster synthesis), `free_lunches_levin` (entry into biology cluster).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification
|
||||
|
||||
- [ ] All 7 deliverables present
|
||||
- [ ] `report.md` 1000-10000 LOC
|
||||
- [ ] Tests pass
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md)
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md)
|
||||
@@ -0,0 +1,91 @@
|
||||
# Track: Video Analysis — Multiscale Logic of Collective Intelligence (Hoffman & Prakash)
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** C (Biological / cognitive / generic systems)
|
||||
|
||||
> **Parent:** Child #10 of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | A Multiscale Logic of Collective Intelligence |
|
||||
| **Author** | Donald Hoffman and Chetan Prakash |
|
||||
| **URL** | https://youtu.be/YnfaT5APPB0 |
|
||||
| **Cluster** | C |
|
||||
| **Slug** | `multiscale_hoffman` |
|
||||
| **Execution order** | #10 of 12 (synthesis across Cluster C) |
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Frames | `artifacts/frames/*.jpg` | 50-500 |
|
||||
| Extraction meta | `artifacts/extraction_meta.json` | Frame paths + hashes |
|
||||
| OCR | `artifacts/ocr.md` | Full OCR per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC** |
|
||||
| Summary | `summary.md` | 200-400 words |
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline
|
||||
|
||||
- [ ] **Phase 1:** Acquire
|
||||
- [ ] **Phase 2:** Keyframes
|
||||
- [ ] **Phase 3:** OCR
|
||||
- [ ] **Phase 4:** Synthesis (1000-10000 LOC)
|
||||
- [ ] **Phase 5:** Verification
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure
|
||||
|
||||
8 sections per umbrella spec §FR6.
|
||||
|
||||
```
|
||||
# A Multiscale Logic of Collective Intelligence
|
||||
**Source:** https://youtu.be/YnfaT5APPB0
|
||||
**Author:** Donald Hoffman and Chetan Prakash
|
||||
**Cluster:** C
|
||||
**Slug:** multiscale_hoffman
|
||||
|
||||
## 1. TL;DR
|
||||
## 2. Key Concepts ← expect conscious agents, multiscale networks, fitness vs truth
|
||||
## 3. Frame Analysis
|
||||
## 4. Transcript Highlights
|
||||
## 5. Mathematical / Theoretical Content
|
||||
## 6. Connections
|
||||
## 7. Open Questions
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections
|
||||
|
||||
- **Forward to:** `cs336_architectures` (collective intelligence in transformers?), `creikey_dl_cv` (applied capstone).
|
||||
- **Backward from:** `neural_dynamics_miller` (concrete neuro), `brain_counterintuitive` (other biological), `generic_systems_fields` (meta-frame).
|
||||
- **Likely rich cross-references:** `free_lunches_levin` (both about collective/Platonic systems), `generic_systems_fields` (both meta-theoretical).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification
|
||||
|
||||
- [ ] All 7 deliverables present
|
||||
- [ ] `report.md` 1000-10000 LOC
|
||||
- [ ] Tests pass
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md)
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md)
|
||||
@@ -0,0 +1,91 @@
|
||||
# Track: Video Analysis — Cognition Emerges from Neural Dynamics (Miller)
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** C (Biological / cognitive / generic systems)
|
||||
|
||||
> **Parent:** Child #9 of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | Cognition Emerges from Neural Dynamics |
|
||||
| **Author** | Earl Miller |
|
||||
| **URL** | https://youtu.be/0BS-BzEFTXA |
|
||||
| **Cluster** | C |
|
||||
| **Slug** | `neural_dynamics_miller` |
|
||||
| **Execution order** | #9 of 12 (concrete neuro) |
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Frames | `artifacts/frames/*.jpg` | 50-500 |
|
||||
| Extraction meta | `artifacts/extraction_meta.json` | Frame paths + hashes |
|
||||
| OCR | `artifacts/ocr.md` | Full OCR per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC** |
|
||||
| Summary | `summary.md` | 200-400 words |
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline
|
||||
|
||||
- [ ] **Phase 1:** Acquire
|
||||
- [ ] **Phase 2:** Keyframes
|
||||
- [ ] **Phase 3:** OCR
|
||||
- [ ] **Phase 4:** Synthesis (1000-10000 LOC)
|
||||
- [ ] **Phase 5:** Verification
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure
|
||||
|
||||
8 sections per umbrella spec §FR6.
|
||||
|
||||
```
|
||||
# Cognition Emerges from Neural Dynamics
|
||||
**Source:** https://youtu.be/0BS-BzEFTXA
|
||||
**Author:** Earl Miller
|
||||
**Cluster:** C
|
||||
**Slug:** neural_dynamics_miller
|
||||
|
||||
## 1. TL;DR
|
||||
## 2. Key Concepts ← expect cortical dynamics, working memory, cognitive flexibility
|
||||
## 3. Frame Analysis
|
||||
## 4. Transcript Highlights
|
||||
## 5. Mathematical / Theoretical Content
|
||||
## 6. Connections
|
||||
## 7. Open Questions
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections
|
||||
|
||||
- **Forward to:** `multiscale_hoffman` (synthesis across scales).
|
||||
- **Backward from:** `brain_counterintuitive` (other concrete neuro), `generic_systems_fields` (meta-frame).
|
||||
- **Likely rich cross-references:** `brain_counterintuitive` (most direct — both about brain mechanisms).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification
|
||||
|
||||
- [ ] All 7 deliverables present
|
||||
- [ ] `report.md` 1000-10000 LOC
|
||||
- [ ] Tests pass
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md)
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md)
|
||||
@@ -0,0 +1,91 @@
|
||||
# Track: Video Analysis — Towards a Platonic Intelligence (Kumar)
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** B (Platonic / geometric AI representations)
|
||||
|
||||
> **Parent:** Child #5 of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | Towards a Platonic Intelligence with Unified Factored Representations |
|
||||
| **Author** | Akarsh Kumar |
|
||||
| **URL** | https://youtu.be/1mXUFweWOug |
|
||||
| **Cluster** | B |
|
||||
| **Slug** | `platonic_intelligence_kumar` |
|
||||
| **Execution order** | #5 of 12 (geometric/Platonic framing) |
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Frames | `artifacts/frames/*.jpg` | 50-500 |
|
||||
| Extraction meta | `artifacts/extraction_meta.json` | Frame paths + hashes |
|
||||
| OCR | `artifacts/ocr.md` | Full OCR per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC** |
|
||||
| Summary | `summary.md` | 200-400 words |
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline
|
||||
|
||||
- [ ] **Phase 1:** Acquire
|
||||
- [ ] **Phase 2:** Keyframes
|
||||
- [ ] **Phase 3:** OCR
|
||||
- [ ] **Phase 4:** Synthesis (1000-10000 LOC)
|
||||
- [ ] **Phase 5:** Verification
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure
|
||||
|
||||
8 sections per umbrella spec §FR6.
|
||||
|
||||
```
|
||||
# Towards a Platonic Intelligence with Unified Factored Representations
|
||||
**Source:** https://youtu.be/1mXUFweWOug
|
||||
**Author:** Akarsh Kumar
|
||||
**Cluster:** B
|
||||
**Slug:** platonic_intelligence_kumar
|
||||
|
||||
## 1. TL;DR
|
||||
## 2. Key Concepts ← expect "Platonic Representation Hypothesis", unified factored reps
|
||||
## 3. Frame Analysis
|
||||
## 4. Transcript Highlights
|
||||
## 5. Mathematical / Theoretical Content ← dominant
|
||||
## 6. Connections
|
||||
## 7. Open Questions
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections
|
||||
|
||||
- **Forward to:** `free_lunches_levin` (Platonic + agential), `cs336_architectures` (Platonic representations in LLMs).
|
||||
- **Backward from:** `score_dynamics_giorgini` (math foundations), `cs229_building_llms` (ML setup).
|
||||
- **Likely rich cross-references:** `free_lunches_levin` (most direct — both about Platonic representations, different angles).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification
|
||||
|
||||
- [ ] All 7 deliverables present
|
||||
- [ ] `report.md` 1000-10000 LOC
|
||||
- [ ] Tests pass
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md)
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md)
|
||||
@@ -0,0 +1,93 @@
|
||||
# Track: Video Analysis — Probability Theory is an Extension of Logic
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** A (Math & information-theoretic foundations)
|
||||
|
||||
> **Parent:** This is child #2 of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella. See [umbrella spec.md](../../video_analysis_campaign_20260621/spec.md) for the full design and multi-pass context.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | Probability Theory is an Extension of Logic |
|
||||
| **Author** | (unknown — verify during execution) |
|
||||
| **URL** | https://youtu.be/0yF9TvMeAzM |
|
||||
| **Cluster** | A (Math & information-theoretic foundations) |
|
||||
| **Slug** | `probability_logic` |
|
||||
| **Execution order** | #2 of 12 (pure math/logic; builds on CS229's ML setup) |
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript (timestamped + plain) | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Extracted unique frames | `artifacts/frames/*.jpg` | 50-500 frames |
|
||||
| Extraction metadata | `artifacts/extraction_meta.json` | Frame paths + hashes + timestamps |
|
||||
| OCR results | `artifacts/ocr.md` | Full OCR text per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC markdown** |
|
||||
| Quick summary | `summary.md` | 200-400 words |
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline (5 phases)
|
||||
|
||||
- [ ] **Phase 1: Acquire.** `extract_transcript.py` + `download_video.py`.
|
||||
- [ ] **Phase 2: Keyframes.** `extract_keyframes.py` with threshold 0.4.
|
||||
- [ ] **Phase 3: OCR.** `ocr_frames.py`.
|
||||
- [ ] **Phase 4: Synthesis.** Tier 3 worker: transcript + OCR + frames → `report.md` (1000-10000 LOC).
|
||||
- [ ] **Phase 5: Verification.** Idempotency + audit + end-of-track report.
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure (8 sections)
|
||||
|
||||
Per umbrella spec §FR6.
|
||||
|
||||
```
|
||||
# Probability Theory is an Extension of Logic
|
||||
**Source:** https://youtu.be/0yF9TvMeAzM
|
||||
**Author:** <verify>
|
||||
**Date Added to Campaign:** 2026-06-21
|
||||
**Cluster:** A
|
||||
**Slug:** probability_logic
|
||||
|
||||
## 1. TL;DR
|
||||
## 2. Key Concepts
|
||||
## 3. Frame Analysis
|
||||
## 4. Transcript Highlights
|
||||
## 5. Mathematical / Theoretical Content ← likely dominant
|
||||
## 6. Connections to Other Videos
|
||||
## 7. Open Questions
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections
|
||||
|
||||
- **Forward to:** `entropy_epiplexity` (info-theoretic framing), `score_dynamics_giorgini` (uses probability), `cs336_architectures` (transformer attention uses probability).
|
||||
- **Backward from:** `cs229_building_llms` (sets canonical ML vocabulary).
|
||||
- **Likely rich cross-references:** `entropy_epiplexity` (most direct — both about extending logic/math).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification
|
||||
|
||||
- [ ] All 7 deliverables present
|
||||
- [ ] `report.md` 1000-10000 LOC
|
||||
- [ ] All 8 sections populated
|
||||
- [ ] Tests pass
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md)
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md)
|
||||
@@ -0,0 +1,91 @@
|
||||
# Track: Video Analysis — Learning Dynamics from Statistics (Giorgini)
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only child track (Pass 1 of 3)
|
||||
**Owner:** Tier 2 Tech Lead (execution)
|
||||
**Cluster:** A (Math & information-theoretic foundations)
|
||||
|
||||
> **Parent:** Child #4 of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella.
|
||||
|
||||
---
|
||||
|
||||
## 1. Video
|
||||
|
||||
| Field | Value |
|
||||
|---|---|
|
||||
| **Title** | Learning Dynamics from Statistics: a score-based approach |
|
||||
| **Author** | Ludovico Giorgini |
|
||||
| **URL** | https://youtu.be/P75iVMmbqQk |
|
||||
| **Cluster** | A |
|
||||
| **Slug** | `score_dynamics_giorgini` |
|
||||
| **Execution order** | #4 of 12 (bridges math → learning theory) |
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Target |
|
||||
|---|---|---|
|
||||
| Transcript | `artifacts/transcript.json` | All segments |
|
||||
| Download log | `artifacts/download.log` | yt-dlp output |
|
||||
| Frames | `artifacts/frames/*.jpg` | 50-500 |
|
||||
| Extraction meta | `artifacts/extraction_meta.json` | Frame paths + hashes |
|
||||
| OCR | `artifacts/ocr.md` | Full OCR per frame |
|
||||
| Deep-dive report | `report.md` | **1000-10000 LOC** |
|
||||
| Summary | `summary.md` | 200-400 words |
|
||||
|
||||
---
|
||||
|
||||
## 3. Pipeline
|
||||
|
||||
- [ ] **Phase 1:** Acquire
|
||||
- [ ] **Phase 2:** Keyframes
|
||||
- [ ] **Phase 3:** OCR
|
||||
- [ ] **Phase 4:** Synthesis (1000-10000 LOC)
|
||||
- [ ] **Phase 5:** Verification
|
||||
|
||||
---
|
||||
|
||||
## 4. Report structure
|
||||
|
||||
8 sections per umbrella spec §FR6.
|
||||
|
||||
```
|
||||
# Learning Dynamics from Statistics: a score-based approach
|
||||
**Source:** https://youtu.be/P75iVMmbqQk
|
||||
**Author:** Ludovico Giorgini
|
||||
**Cluster:** A
|
||||
**Slug:** score_dynamics_giorgini
|
||||
|
||||
## 1. TL;DR
|
||||
## 2. Key Concepts ← expect score-based generative models, score matching, SDEs
|
||||
## 3. Frame Analysis
|
||||
## 4. Transcript Highlights
|
||||
## 5. Mathematical / Theoretical Content ← dominant
|
||||
## 6. Connections
|
||||
## 7. Open Questions
|
||||
## 8. References
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. Connections
|
||||
|
||||
- **Forward to:** `cs336_architectures` (score-based connections to diffusion LLMs? if applicable), `creikey_dl_cv` (applied score-based models).
|
||||
- **Backward from:** `entropy_epiplexity` (model complexity), `probability_logic` (probability foundations).
|
||||
- **Likely rich cross-references:** `entropy_epiplexity` (model complexity informs training dynamics).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification
|
||||
|
||||
- [ ] All 7 deliverables present
|
||||
- [ ] `report.md` 1000-10000 LOC
|
||||
- [ ] Tests pass
|
||||
|
||||
---
|
||||
|
||||
## 7. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md)
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md)
|
||||
@@ -0,0 +1,125 @@
|
||||
# Track: Video Analysis Campaign — Synthesis (2026-06-21)
|
||||
|
||||
**Status:** Not started (umbrella published 2026-06-21)
|
||||
**Type:** Research-only synthesis track (Pass 1 of 3)
|
||||
**Owner:** Tier 1 Orchestrator (synthesis spec + report); Tier 2 Tech Lead (execution)
|
||||
**Priority:** A (delivers the user's requested "summary of each video" + "summary report of key takeaways")
|
||||
**Domain:** Meta-tooling (cross-cutting research synthesis; no `src/` changes)
|
||||
|
||||
> **Parent:** This synthesis track is blocked_by all 12 child tracks of the [video_analysis_campaign_20260621](../../video_analysis_campaign_20260621/) umbrella. See [umbrella spec.md](../../video_analysis_campaign_20260621/spec.md) for the full campaign design and multi-pass context.
|
||||
|
||||
> **Multi-pass note:** The synthesis `report.md` is intermediate input to Pass 2 (de-obfuscation). Per the campaign's lossless-preservation directive (umbrella spec §0), the synthesis must preserve detail — Pass 2 will compress, not this pass.
|
||||
|
||||
---
|
||||
|
||||
## 1. Inputs
|
||||
|
||||
This track consumes the outputs of all 12 child tracks:
|
||||
|
||||
| # | Slug | Cluster | Source |
|
||||
|---|---|---|---|
|
||||
| 1 | `cs229_building_llms` | E | [video_analysis_cs229_building_llms_20260621/](../../video_analysis_cs229_building_llms_20260621/) |
|
||||
| 2 | `probability_logic` | A | [video_analysis_probability_logic_20260621/](../../video_analysis_probability_logic_20260621/) |
|
||||
| 3 | `entropy_epiplexity` | A | [video_analysis_entropy_epiplexity_20260621/](../../video_analysis_entropy_epiplexity_20260621/) |
|
||||
| 4 | `score_dynamics_giorgini` | A | [video_analysis_score_dynamics_giorgini_20260621/](../../video_analysis_score_dynamics_giorgini_20260621/) |
|
||||
| 5 | `platonic_intelligence_kumar` | B | [video_analysis_platonic_intelligence_kumar_20260621/](../../video_analysis_platonic_intelligence_kumar_20260621/) |
|
||||
| 6 | `free_lunches_levin` | B | [video_analysis_free_lunches_levin_20260621/](../../video_analysis_free_lunches_levin_20260621/) |
|
||||
| 7 | `generic_systems_fields` | C | [video_analysis_generic_systems_fields_20260621/](../../video_analysis_generic_systems_fields_20260621/) |
|
||||
| 8 | `brain_counterintuitive` | C | [video_analysis_brain_counterintuitive_20260621/](../../video_analysis_brain_counterintuitive_20260621/) |
|
||||
| 9 | `neural_dynamics_miller` | C | [video_analysis_neural_dynamics_miller_20260621/](../../video_analysis_neural_dynamics_miller_20260621/) |
|
||||
| 10 | `multiscale_hoffman` | C | [video_analysis_multiscale_hoffman_20260621/](../../video_analysis_multiscale_hoffman_20260621/) |
|
||||
| 11 | `cs336_architectures` | E | [video_analysis_cs336_architectures_20260621/](../../video_analysis_cs336_architectures_20260621/) |
|
||||
| 12 | `creikey_dl_cv` | D | [video_analysis_creikey_dl_cv_20260621/](../../video_analysis_creikey_dl_cv_20260621/) |
|
||||
|
||||
**Per-child inputs consumed:**
|
||||
- `report.md` (the 1000-10000 LOC deep-dive)
|
||||
- `summary.md` (the 200-400 word quick summary)
|
||||
|
||||
The per-child `transcript.json`, `frames/`, `ocr.md`, and `extraction_meta.json` are NOT consumed here — they feed Pass 2 directly.
|
||||
|
||||
---
|
||||
|
||||
## 2. Deliverables
|
||||
|
||||
| Artifact | Path | Description |
|
||||
|---|---|---|
|
||||
| Per-video roll-up | `per_video_summary.md` | One paragraph (150-250 words) per video — the "summary of each video" the user requested. Ordered by execution order (matches umbrella §6). |
|
||||
| Synthesis report | `report.md` | The "summary report of key takeaways" — 6 sections per umbrella §FR7. |
|
||||
|
||||
---
|
||||
|
||||
## 3. Synthesis report structure (6 sections)
|
||||
|
||||
Per umbrella spec §FR7.
|
||||
|
||||
```
|
||||
# Video Analysis Campaign — Synthesis
|
||||
|
||||
## 1. Theme Matrix (across clusters A/B/C/D/E)
|
||||
## 2. Cross-Video Concept Map
|
||||
## 3. 5-10 High-Level Takeaways
|
||||
## 4. Mathematical Prerequisite Graph
|
||||
## 5. Open Research Questions
|
||||
## 6. Recommended Next-Watch List
|
||||
```
|
||||
|
||||
**Section detail:**
|
||||
|
||||
**§1 Theme Matrix** — a 2D table with rows = clusters (A/B/C/D/E) and columns = themes (e.g., foundations, representations, training, applications, biological inspiration, ethics). Each cell: which videos address this theme.
|
||||
|
||||
**§2 Cross-Video Concept Map** — for each major concept that appeared in 2+ videos, list: (a) which videos introduced it, (b) which videos built on it, (c) which videos referenced it. Format: per-concept subsection with a list of video slugs + brief role description.
|
||||
|
||||
**§3 5-10 High-Level Takeaways** — bullet list of the most important cross-cutting insights the user should walk away with. Each takeaway: 2-5 sentences with references to the videos that support it.
|
||||
|
||||
**§4 Mathematical Prerequisite Graph** — a directed graph showing which mathematical concepts are needed to understand which. E.g., "to understand CS336 Lecture 3, you need: linear algebra (CS229), probability (Probability = Extension of Logic), score-based dynamics (Giorgini)." Format: text-based DAG or ASCII graph.
|
||||
|
||||
**§5 Open Research Questions** — questions raised by the videos that the field doesn't have consensus answers to. The user mentioned this is "largely impenetrable" to them and associates — these questions are the campaign's open frontier.
|
||||
|
||||
**§6 Recommended Next-Watch List** — based on what the user liked in this batch, suggest related videos/authors/topics to investigate next. Source from cross-references in the per-video reports + the user's stated stylistic preferences.
|
||||
|
||||
---
|
||||
|
||||
## 4. Pipeline (per umbrella spec §FR7)
|
||||
|
||||
- [ ] **Phase 1: Ingest.** Read all 12 child `report.md` files + `summary.md` files. Build an in-memory index.
|
||||
- [ ] **Phase 2: Per-video roll-up.** Generate `per_video_summary.md` by either lifting each child's `summary.md` (preferred) or writing a 150-250 word summary if the child's is too short.
|
||||
- [ ] **Phase 3: Synthesis report.** Generate `report.md` per the 6-section structure above. Pass 1 of 3 = detailed; Pass 2 will compress.
|
||||
- [ ] **Phase 4: Verification.** Cross-check that every video in the campaign has a roll-up entry. Cross-check that every theme in §1 is sourced from at least one video. Cross-check that every takeaway in §3 has at least one supporting video reference.
|
||||
|
||||
---
|
||||
|
||||
## 5. Lossless preservation directive
|
||||
|
||||
Per umbrella spec §0: this synthesis is intermediate input to Pass 2 (de-obfuscation). DO NOT over-summarize. The §3 takeaways should be 2-5 sentences each (not 1 sentence). The §4 math prerequisite graph should reference specific videos, not just "foundational math." The §5 open research questions should include the user's own context (what's impenetrable to them) and not just generic AI debates.
|
||||
|
||||
If the synthesis report is less than 1000 LOC, it is too short. Target: 1000-5000 LOC for the synthesis report (less than per-video because the heavy lifting is in the per-video reports).
|
||||
|
||||
---
|
||||
|
||||
## 6. Verification (per umbrella spec §12)
|
||||
|
||||
- [ ] All 12 children shipped (read all their `report.md` + `summary.md`)
|
||||
- [ ] `per_video_summary.md` has 12 entries (one per video), each 150-250 words
|
||||
- [ ] `report.md` has all 6 sections populated
|
||||
- [ ] `report.md` is 1000-5000 LOC (detailed enough for Pass 2 to compress)
|
||||
- [ ] Every §3 takeaway references at least one video
|
||||
- [ ] Every §1 theme cell references at least one video
|
||||
- [ ] §6 next-watch list references at least 3 sources
|
||||
|
||||
---
|
||||
|
||||
## 7. Out of scope (per umbrella spec §10)
|
||||
|
||||
- De-obfuscation (Pass 2 — future track, user must first rediscover encoding notation)
|
||||
- Projection to applied domain (Pass 3 — future track, user must first articulate "own caveats")
|
||||
- Modifying any `src/*.py` files in manual_slop
|
||||
- Building a video knowledge base (separate dimension per `conductor/code_styleguides/knowledge_artifacts.md`)
|
||||
|
||||
---
|
||||
|
||||
## 8. See also
|
||||
|
||||
- [Umbrella spec.md](../../video_analysis_campaign_20260621/spec.md) — full campaign design + multi-pass context
|
||||
- [Umbrella plan.md](../../video_analysis_campaign_20260621/plan.md) — campaign-level plan
|
||||
- [Umbrella README.md](../../video_analysis_campaign_20260621/README.md) — child index
|
||||
- All 12 child `spec.md` files (linked in §1 above)
|
||||
Reference in New Issue
Block a user