Challenge: Recent studies show that supervised fine-tuning (SFT) is a common approach for reasoning in large language models.
Approach: They propose to use supervised fine-tuning (SFT) on chain-of-thought trajectories demonstrations . they find that incorporating negative traxories yields substantial OOD generalization gains .
Outcome: The proposed scheme yields 5.51% OOD gain over positive-only training.

Similar Papers

NAT: Enhancing Agent Tuning with Negative Samples (2025.naacl-long)

Copied to clipboard

Challenge: Existing methods for fine-tuning and reinforcement learning use only positive examples, limiting their efficiency in low-resource scenarios.
Approach: They propose a method that leverages both successful and failed trajectories for fine-tuning, maximizing the utility of limited resources.
Outcome: The proposed method surpasses existing methods, including SFT, DPO, and PPO, across various tasks.
Can LLMs Learn From Mistakes? An Empirical Study on Reasoning Tasks (2024.findings-emnlp)

Copied to clipboard

Challenge: Existing work has shown that simple learning can enhance the chain-of-thought (CoT) reasoning of large language models.
Approach: They construct mistake-correction datasets to identify and correct mistakes in CoTs . they conclude that LLMs can learn from mistakes to enhance their CoT reasoning .
Outcome: The proposed datasets show that LLMs can learn from mistakes to enhance their CoT reasoning performance.
Unearthing Gems from Stones: Policy Optimization with Negative Sample Augmentation for LLM Reasoning (2025.findings-emnlp)

Copied to clipboard

Challenge: Recent advances in reasoning language models have witnessed a paradigm shift from short to long CoT pattern.
Approach: They propose a behavior-constrained policy gradient with negative sample augmented (BCPG-NSA) negative steps are valuable components in long CoT models, authors argue .
Outcome: The proposed framework outperforms baselines on math/coding reasoning benchmarks using the same training dataset.
VCORE: Variance-Controlled Optimization-based Reweighting for Chain-of-Thought Supervision (2026.acl-long)

Copied to clipboard

Challenge: Empirical evaluations demonstrate that VCORE achieves the strongest overall average performance, with especially clear gains on lower-capacity models.
Approach: They propose a framework that reformulates supervision as a constrained optimization problem.
Outcome: Empirical evaluations show that VCORE achieves the strongest overall average performance, with especially clear gains on lower-capacity models.
Turning Failures into Value: Negative Experience Replay for RLVR via Confidence Gating and Boundary Failure Sampling (2026.acl-long)

Copied to clipboard

Challenge: Existing experience replay methods for RLVR ignore sample inefficiency . expensive reasoning trajectories are discarded immediately after a single gradient update .
Approach: They propose a method to replay failure trajectories to improve model refinement . they propose 'nexGRPO' which employs mid-confidence gating to filter invalid noise and saturated errors.
Outcome: The proposed model outperforms strong baaselines and achieves improved out-of-distribution generalization.
Beyond Rejection Sampling: Trajectory Fusion for Scaling Mathematical Reasoning (2026.findings-acl)

Copied to clipboard

Challenge: Large language models (LLMs) fine-tuned using rejection sampling retain only correct reasoning trajectories . however, this paradigm treats supervision as a binary filter that systematically excludes teacher-generated errors, leaving a gap in how reasoning failures are modeled during training.
Approach: They propose a fine-tuning strategy that reframes rejection sampling as a structured supervision construction process.
Outcome: The proposed approach outperforms RFT on multiple math benchmarks while retaining only correct reasoning trajectories.
ReFT: Reasoning with Reinforced Fine-Tuning (2024.acl-long)

Copied to clipboard

Challenge: Existing approaches to improve the generalization of large language models are using Supervised Fine-Tuning (SFT) this approach does not show sufficient generalization ability because it only relies on the given CoT data.
Approach: They propose to use Chain-of-Thought annotations to train Large Language Models using supervised fine-tuning to improve generalization.
Outcome: The proposed approach outperforms SFT on GSM8K, MathQA, and SVAMP datasets and shows a superior generalization ability.
Anti-Length Shift: Dynamic Outlier Truncation for Training Efficient Reasoning Models (2026.acl-long)

Copied to clipboard

Challenge: Existing efficient reasoning methods rely on explicit length penalties for excessive verbosity on simple queries.
Approach: They propose a training-time intervention that selectively suppresses redundant tokens . they find length shift occurs when models generate unnecessary reasoning on trivial inputs - a phenomenon that is often unexplored .
Outcome: The proposed method reduces inference token usage by 78% while increasing accuracy compared to the initial policy and surpasses state-of-the-art efficient reasoning methods.
Language Models Can Easily Learn to Reason from Demonstrations (2025.findings-emnlp)

Copied to clipboard

Challenge: Large reasoning models (LRMs) tackle complex problems by following long chain-of-thoughts (Long CoT) however, the training techniques and data requirements to elicit Long CoT remain poorly understood.
Approach: They propose to use data-efficient supervised fine-tuning and parameter-efficient low-rank adaptation to elicit Long CoT reasoning.
Outcome: The proposed model can learn Long CoT reasoning through data-efficient supervised fine-tuning and parameter-efficient low-rank adaptation.
ProFit: Leveraging High-Value Signals in SFT via Probability-Guided Token Selection (2026.findings-acl)

Copied to clipboard

Challenge: Traditional fine-tuning ignores one-to-many nature of language, leading to overfitting . authors propose a method to fine- tune LLMs by leveraging tokens.
Approach: They propose a method to fine-tune Large Language Models by leveraging tokens to mask low-probability tokens.
Outcome: The proposed method outperforms baselines on general reasoning and mathematical benchmarks.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations