Innovation Prior — reconstructed reasoning traces
bohanlyu.com · 232 words · saved by 1 readers
Reconstructed first-person reasoning traces for famous and recent papers: re-deriving each method from the pain points and prior art of its time.
Innovation Prior - reconstructed reasoning traces Skip to content Innovation Prior Every paper here is rewritten as a first-person discovery trace - the method re-derived from the pain points and prior art of its time, hitting walls and self-correcting until it lands on the idea and the code. Finished papers are written theorem-first: the answer arrives polished, with the struggle edited out. Innovation Prior runs that backwards. For each paper we reconstruct the chain of thought of how the method was actually arrived at, starting from the state of the field before it existed. Each method come
saved by
related reading
- Circuit Tracing: Revealing Computational Graphs in Language Modelstransformer-circuits.pub
- Automated Weak-to-Strong Researcheralignment.anthropic.com
- What is First Principles Thinking?fs.blog
- DeepSeek-R1arxiv.org
- Stolen Thoughtsstolen-thoughts.com
- [2603.07267] How to Steal Reasoning Without Reasoning Tracesarxiv.org
- [2203.14465] STaR: Bootstrapping Reasoning With Reasoningarxiv.org
- Reasoning as Trajectoriesslhleosun.github.io
- Expert Data for Frontier AI - AfterQueryafterquery.com
- [2608.20319] Inducing Task Models from Computer-Use Tracesarxiv.org
- Paper Trailspapertrailshq.com
- Stealing Reasoning Traces from Proprietary LLM APIsresearch.snyk.io