Why Do Reasoning Models Lose Coverage? The Role of Data and Forks in the Road

• URL: http://arxiv.org/abs/2605.17026v1
• Date: Sat, 16 May 2026 14:55:12 GMT
• Title: Why Do Reasoning Models Lose Coverage? The Role of Data and Forks in the Road
• Abstract summary: We investigate the reasoning shrinkage arise under SFT-based post-training.<n>By tracking post-training dynamics in settings, we find that the shrinkage phenomenon is tightly correlated with the prevalence of decision-point scenarios in the training data.<n>Our findings identify data-centric factors as a key driver of shrinkage in reasoning models and highlight diversity-aware designs as an effective lever for controlling it.
• Score: 26.088108937790167
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Recent progress in large language models has led to the emergence of reasoning models, which have shown strong performance on complex tasks through specialized fine-tuning procedures. While these methods reliably improve pass@1 accuracy, prior works have observed that they show a coverage shrinkage behavior, where pass@k degrades relative to the base model. In this paper, we investigate the reasoning shrinkage arise under SFT-based post-training. We hypothesize that this behavior is driven by properties of the fine-tuning data, specifically related to decision points or "forks in the road" scenarios where model faces indecipherable patterns with multiple valid reasoning paths. To test this hypothesis, we design controlled case studies that simulate such decision-point settings, spanning indecipherable nodes in graph branching, and reasoning modes. By tracking post-training dynamics in these settings, we find that the shrinkage phenomenon is tightly correlated with the prevalence of decision-point scenarios in the training data. We also demonstrate that this shrinkage behavior can be partially mitigated through targeted data synthesis design of decision-points, and a more systematic diversity-encouraging decoding mechanism. Our findings identify data-centric factors as a key driver of shrinkage in reasoning models and highlight diversity-aware designs as an effective lever for controlling it.

Related papers

• CausalVAD: De-confounding End-to-End Autonomous Driving via Causal Intervention [24.03101162891974]
  Planning-oriented end-to-end driving models learn statistical correlations instead of true causal relationships.<n>This vulnerability leads to causal confusion, where models exploit dataset biases as shortcuts.<n>We introduce CausalVAD, a de-confounding training framework that leverages causal intervention.
  arXiv  Detail & Related papers  (2026-03-19T07:21:01Z)
• Decoding the Critique Mechanism in Large Reasoning Models [50.821607345799386]
  Large Reasoning Models (LRMs) exhibit backtracking and self-verification mechanisms that enable them to revise intermediate steps and reach correct solutions.<n>We investigate how current LRMs recover from errors by inserting arithmetic mistakes in their intermediate reasoning steps.<n>We find that despite the error propagating through the chain-of-thought, the model still reaches the correct final answer.
  arXiv  Detail & Related papers  (2026-03-17T10:03:30Z)
• InjectRBP: Steering Large Language Model Reasoning Behavior via Pattern Injection [12.65760977924031]
  Reasoning can significantly enhance the performance of Large Language Models.<n>We investigate how models' reasoning behaviors shape reasoning from the perspective of behavioral patterns.<n>We propose two optimization methods that require no parameter updates: InjectCorrect and InjectRLOpt.
  arXiv  Detail & Related papers  (2026-02-12T14:44:40Z)
• Preference Learning for AI Alignment: a Causal Perspective [55.2480439325792]
  We frame this problem in a causal paradigm, providing the rich toolbox of causality to identify persistent challenges.<n>Inheriting from the literature of causal inference, we identify key assumptions necessary for reliable generalisation.<n>We illustrate failure modes of naive reward models and demonstrate how causally-inspired approaches can improve model robustness.
  arXiv  Detail & Related papers  (2025-06-06T10:45:42Z)
• The CoT Encyclopedia: Analyzing, Predicting, and Controlling how a Reasoning Model will Think [81.38614558541772]
  We introduce the CoT Encyclopedia, a framework for analyzing and steering model reasoning.<n>Our method automatically extracts diverse reasoning criteria from model-generated CoTs.<n>We show that this framework produces more interpretable and comprehensive analyses than existing methods.
  arXiv  Detail & Related papers  (2025-05-15T11:31:02Z)
• Attuned to Change: Causal Fine-Tuning under Latent-Confounded Shifts [32.989526411946606]
  Adapting to latent-confounded shifts remains a core challenge in modern AI.<n>One practical failure mode arises when fine-tuning pre-trained foundation models on confounded data.<n>We frame causal fine-tuning as an identification problem and pose an explicit causal model that decomposes inputs into low-level spurious features.
  arXiv  Detail & Related papers  (2024-10-18T11:06:23Z)
• Influence Functions for Scalable Data Attribution in Diffusion Models [52.92223039302037]
  Diffusion models have led to significant advancements in generative modelling.<n>Yet their widespread adoption poses challenges regarding data attribution and interpretability.<n>We develop an influence functions framework to address these challenges.
  arXiv  Detail & Related papers  (2024-10-17T17:59:02Z)
• Causal Analysis for Robust Interpretability of Neural Networks [0.2519906683279152]
  We develop a robust interventional-based method to capture cause-effect mechanisms in pre-trained neural networks. We apply our method to vision models trained on classification tasks.
  arXiv  Detail & Related papers  (2023-05-15T18:37:24Z)
• Revisiting Design Choices in Model-Based Offline Reinforcement Learning [39.01805509055988]
  Offline reinforcement learning enables agents to leverage large pre-collected datasets of environment transitions to learn control policies. This paper compares and designs novel protocols to investigate their interaction with other hyper parameters, such as the number of models, or imaginary rollout horizon.
  arXiv  Detail & Related papers  (2021-10-08T13:51:34Z)
• Variational Causal Networks: Approximate Bayesian Inference over Causal Structures [132.74509389517203]
  We introduce a parametric variational family modelled by an autoregressive distribution over the space of discrete DAGs. In experiments, we demonstrate that the proposed variational posterior is able to provide a good approximation of the true posterior.
  arXiv  Detail & Related papers  (2021-06-14T17:52:49Z)
• Structural Causal Models Are (Solvable by) Credal Networks [70.45873402967297]
  Causal inferences can be obtained by standard algorithms for the updating of credal nets. This contribution should be regarded as a systematic approach to represent structural causal models by credal networks. Experiments show that approximate algorithms for credal networks can immediately be used to do causal inference in real-size problems.
  arXiv  Detail & Related papers  (2020-08-02T11:19:36Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.