The paper investigates why fine‑tuned reasoning models lose coverage, observing that pass@k accuracy degrades relative to the base model. The authors attribute this shrinkage to decision‑point or “forks in the road” scenarios in the fine‑tuning data, where the model faces multiple valid reasoning paths. Controlled experiments confirm a strong correlation between such decision‑point prevalence and coverage loss, and show that targeted data synthesis and diversity‑encouraging decoding can partially mitigate the effect.
arXiv:2604. 10827v2 Announce Type: replace Abstract: Compute scaling for LLM reasoning trades off exploring solution approaches (\emph{breadth}) against refining promising ones (\emph{depth}), yet why a given trade-off works, and why it often fails to transfer across models, remains unclear.
arXiv:2608.30413v1 Announce Type: new Abstract: Defeasible reasoning is a type of reasoning where inferences are drawn from plausible current evidence, but can be retracted upon the introduction of n...
Defeasible reasoning is a type of reasoning where inferences are drawn from plausible current evidence, but can be retracted upon the introduction of newer evidence. Although recent studies have exami...
Scientific reasoning models for biology combine language models with foundation models trained on multimodal biological data, including DNA, RNA, and proteins. These models are built through post-training, yet how each stage shapes reasoning and generalization remains poorly understood.
Supervised fine-tuning (SFT) on a small, high-quality set of long reasoning traces is an effective approach for eliciting strong reasoning capabilities in Large Language Models (LLMs). However, existing methods for curating high-quality SFT data rely heavily on strong reasoning models to filter examples based on diversity and difficulty, making the curation process costly while often yielding suboptimal data quality.
arXiv:2607. 20500v1 Announce Type: new Abstract: Large Language Models (LLMs) perform strongly on well-specified reasoning tasks with a feasible answer.
The paper investigates Evolution Strategies (ES) as a memory‑efficient post‑training method for large language model (LLM) reasoning. It demonstrates that ES outperforms Group Relative Policy Optimization (GRPO) by achieving broader reasoning coverage, improving Pass@K metrics, and avoiding entropy collapse. The study also reveals that ES’s performance gains stem from sparse, high‑magnitude parameter updates, do not cause catastrophic forgetting, and can be combined with GRPO in a sequential training strategy.
arXiv:2604. 06628v2 Announce Type: replace Abstract: A prevailing narrative in LLM post-training holds that supervised finetuning (SFT) memorizes while reinforcement learning (RL) generalizes.
arXiv:2606. 26797v1 Announce Type: new Abstract: Supervised fine-tuning (SFT) on a small, high-quality set of long reasoning traces is an effective approach for eliciting strong reasoning capabilities in Large Language Models (LLMs).
arXiv:2606. 07720v1 Announce Type: new Abstract: Large language models (LLMs) have demonstrated remarkable reasoning abilities on mathematical and multi-hop planning tasks.
arXiv:2606. 16517v1 Announce Type: new Abstract: Scientific reasoning models for biology combine language models with foundation models trained on multimodal biological data, including DNA, RNA, and proteins.
arXiv:2505. 20161v2 Announce Type: replace-cross Abstract: Effective generalization in language models depends critically on the diversity of their training data.