Results and benchmarks
Let LLMs Break Free from Overthinking via Self-Braking Tuning is the primary contribution described in this paper.
| Task | Dataset | Metric | Value | Source |
|---|---|---|---|---|
| Self-braking Tuning | GSM8K | Accuracy | 0.1 | paper-derived |
| Self-braking Tuning | Baseline | Accuracy | 59.36 | paper-derived |
| Self-braking Tuning | SBT-Exact | Accuracy | 57.83 | paper-derived |
| Self-braking Tuning | SBT-E | Accuracy | 56.48 | paper-derived |
Audit each benchmark finding before selecting an implementation path. Evidence refs map to the disclosure below.
Evidence graph: 4 refs, 4 links.
Utility signals: depth 90/100, grounding 95/100, status high.
Implementation
Best maintained implementation now
[NeurIPS 2025] Let LRMs Break Free from Overthinking via Self-Braking Tuning. https://arxiv.org/abs/2505.14604
56 stars · 0 forks · Last push Nov 4, 2025 · Apache-2.0 license
- License
- CI
- Dependencies
- Docker
Official implementation from Papers with Code · Repository link is mentioned in the paper metadata · Strong overlap with paper title keywords
ccai-lab/self-braking-tuning is the strongest maintained implementation based on ranking signals. License is declared (Apache-2.0). Dependency/environment manifests are present.
Open ccai-lab/self-braking-tuning- No CI workflows detected
- Selected ccai-lab/self-braking-tuning as the strongest maintained implementation for new work.
- Includes dependency/environment manifest signals.
- Repository activity is within the last 24 months.
- Official repository is preserved separately as historical context.
Compare implementation paths
Compare maintenance quality, reproducibility coverage, and evidence confidence before choosing a reproduction baseline.
- Maintenance
- Stale risk
- Confidence
- High
- Reproducibility
- Moderate
- Stars
- 56
- Last push
- Nov 4, 2025 (296d)
Official implementation from Papers with Code · Repository link is mentioned in the paper metadata
- No CI pipeline detected
- No tagged releases
- No Docker setup
- Maintenance
- Stale risk
- Confidence
- High
- Reproducibility
- Moderate
- Stars
- 56
- Last push
- Nov 4, 2025 (296d)
Official implementation from Papers with Code · Repository link is mentioned in the paper metadata
- No CI pipeline detected
- No tagged releases
- No Docker setup
Reproduction readiness
Setup required
Dependencies pinned, manual setup needed
- ccai-lab/self-braking-tuning has requirements.txt but requires manual environment setup.
- Last push was 296 days ago, so expect possible dependency version conflicts.
- No Dockerfile, so you will set up the environment manually.
- No CI pipeline, so test coverage is unknown.
Quick start
git clone https://github.com/ccai-lab/self-braking-tuning.git
pip install -r requirements.txtHugging Face artifacts
No direct paper-linked artifacts were found. Showing strongest curated related artifacts for faster exploration.
Models
- KnutJaegersberg/2-bit-LLMs
1,632 downloads · 92 likes
- LLMs/WizardLM-30B-V1.0
112 downloads · 3 likes
- LLMs/WizardLM-13B-V1.0
59 downloads · 4 likes
Broaden model search
Datasets
No trustworthy datasets matches right now.
Search datasets on Hugging FaceSpaces
No trustworthy spaces matches right now.
Search spaces on Hugging FaceResearch context
Tasks
Self-braking Tuning
Methods
None detected
Domains
None detected
Open this paper in HFEPX to review benchmark signals, evaluation modes, and human-feedback protocol context.
Open in HFEPXJump to Paper2Code search queries derived from this paper's research context.
Data includes links from Papers with Code ( CC-BY-SA-4.0 ).