# r2m-ai/load-bearing-cot

<https://repo.masters-masters.com/en/repos/r2m-ai/load-bearing-cot>

[Home](https://repo.masters-masters.com/en) › [r2m-ai](https://repo.masters-masters.com/en/owners/r2m-ai) › [load-bearing-cot](https://repo.masters-masters.com/en/repos/r2m-ai/load-bearing-cot)

r2m-ai/load-bearing-cot has 198 stars on GitHub. Its RepoMasters card: Prototype tier. Perturb-and-continue causal tests of chain-of-thought: does the written…

| Field | Value |
| --- | --- |
| Owner | [r2m-ai](https://repo.masters-masters.com/en/owners/r2m-ai) |
| Stars | 198 |
| Forks | 26 |
| Language | [Python](https://repo.masters-masters.com/en/languages/python) |
| License | MIT |
| Created | 2026 |
| Tier | [Prototype (100 to 249 ★)](https://repo.masters-masters.com/en/ranks/rush) |
| View on GitHub | [r2m-ai/load-bearing-cot](https://github.com/r2m-ai/load-bearing-cot) |

## Repository

Perturb-and-continue causal tests of chain-of-thought: does the written reasoning actually constrain the answer? Code, data, and probes for "From Decorative to Load-Bearing: Task Difficulty Shapes the Causal Role of Chain-of-Thought" (TMLR 2026).

The tier for repos with 100 to 249 ★. Published odds per card pulled: 70%. Rarity: Common.

## Nearby in Python

- [lusouldepth-ai/FDE\_Learning · ★ 198 · Python · Prototype](https://repo.masters-masters.com/en/repos/lusouldepth-ai/FDE_Learning)
- [NewBeeM4n/codex-overload-fix · ★ 198 · Python · Prototype](https://repo.masters-masters.com/en/repos/NewBeeM4n/codex-overload-fix)
- [test-bit-sys/algeria-geo · ★ 198 · Python · Prototype](https://repo.masters-masters.com/en/repos/test-bit-sys/algeria-geo)
- [ask/carrot · ★ 198 · Python · Prototype](https://repo.masters-masters.com/en/repos/ask/carrot)
- [jamesturk/django-markupfield · ★ 198 · Python · Prototype](https://repo.masters-masters.com/en/repos/jamesturk/django-markupfield)
- [louissobel/Drapache · ★ 198 · Python · Prototype](https://repo.masters-masters.com/en/repos/louissobel/Drapache)

---
Data: GitHub · Illustration drawn by RepoMasters from the repo’s ID and its language color.