The reviewed record of science sign in
Pith

arxiv: 2501.10799 · v1 · pith:4FN4IO3E · submitted 2025-01-18 · cs.LG · cs.AI

Step-KTO: Optimizing Mathematical Reasoning through Stepwise Binary Feedback

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:4FN4IO3Erecord.jsonopen to challenge →

classification cs.LG cs.AI
keywords reasoningstep-ktobinaryfeedbackfinalmathematicalaccuracyanswer
0
0 comments X
read the original abstract

Large language models (LLMs) have recently demonstrated remarkable success in mathematical reasoning. Despite progress in methods like chain-of-thought prompting and self-consistency sampling, these advances often focus on final correctness without ensuring that the underlying reasoning process is coherent and reliable. This paper introduces Step-KTO, a training framework that combines process-level and outcome-level binary feedback to guide LLMs toward more trustworthy reasoning trajectories. By providing binary evaluations for both the intermediate reasoning steps and the final answer, Step-KTO encourages the model to adhere to logical progressions rather than relying on superficial shortcuts. Our experiments on challenging mathematical benchmarks show that Step-KTO significantly improves both final answer accuracy and the quality of intermediate reasoning steps. For example, on the MATH-500 dataset, Step-KTO achieves a notable improvement in Pass@1 accuracy over strong baselines. These results highlight the promise of integrating stepwise process feedback into LLM training, paving the way toward more interpretable and dependable reasoning capabilities.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. When Model Merging Breaks Routing: Training-Free Calibration for MoE

    cs.LG 2026-06 unverdicted novelty 6.0

    Merging breaks MoE routing via softmax sensitivity; HARC uses Hessian curvature for closed-form router calibration that improves merged model performance without retraining.

  2. Reason, Reward, Refine: Step-Level Errors Corrections with Structured Feedback for Physics Reasoning in Small Language Models

    cs.AI 2026-07 conditional novelty 5.0

    A step-level reward framework using GPT-4o as a training-time verifier reduces reasoning errors in small language models on physics benchmarks by 10-20% over baselines.

  3. Bridging the Detection-to-Abstention Gap in Reasoning Models under Insufficient Information

    cs.AI 2026-05 unverdicted novelty 5.0

    JTS trains reasoning models via supervised warm-up and missing-premise RL to make an explicit answerability commitment that triggers early termination on unanswerable inputs, raising Abstention@Detection near saturation.