Reasoning Depth and Environment Complexity: A Controlled Study of RLVR Data Allocation across Logical Reasoning Tasks

• URL: http://arxiv.org/abs/2605.26934v1
• Date: Tue, 26 May 2026 12:28:08 GMT
• Title: Reasoning Depth and Environment Complexity: A Controlled Study of RLVR Data Allocation across Logical Reasoning Tasks
• Abstract summary: Reinforcement learning with verifiable rewards (RLVR) has become central to post-training reasoning models.<n>We characterize the reasoning space along two dimensions.<n>We consider four abilities core to real-world reasoning: deductive state tracking, abductive recovery of hidden events or facts, inductive rule induction, and analogical transfer.
• Score: 43.28656684175735
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Reinforcement learning with verifiable rewards (RLVR) has become central to post-training reasoning models, yet a key limitation of existing studies is their narrow view of the reasoning space: difficulty is treated as reasoning depth alone, and reward is concentrated on forward deductive state tracking. We instead characterize the reasoning space along two dimensions. Difficulty. Beyond reasoning depth, we study environment complexity, where models must identify the correct path amid distractors and interacting structures. Rewarded reasoning form. We consider four abilities core to real-world reasoning: deductive state tracking, abductive recovery of hidden events or facts, inductive rule induction, and analogical transfer. To disentangle these factors, we construct a synthetic knowledge-graph environment with controlled pre- and post-training distributions, where each instance varies along depth, complexity, and task family. Three findings emerge: joint depth-complexity coverage outperforms single-axis recipes; reasoning families respond non-uniformly, with abductive reasoning degrading outside the RL-covered region and task correlations clustering into deductive-abductive and inductive-analogy pairs; and uniform mixing outperforms staged curricula under a fixed budget. We also find that recent off-the-shelf models exhibit the same deductive-over-abductive asymmetry, suggesting that this gap is not merely an artifact of our controlled setup.

Related papers

• Compliance versus Sensibility: On the Reasoning Controllability in Large Language Models [46.26628756478016]
  Large Language Models (LLMs) are known to acquire reasoning capabilities through shared inference patterns in pre-training data.<n>We show that reasoning conflicts are internally detectable, as confidence scores significantly drop during conflicting episodes.<n>Our findings establish that while LLM reasoning is anchored to concrete instances, active mechanistic interventions can effectively decouple logical schemata from data.
  arXiv  Detail & Related papers  (2026-04-29T22:55:40Z)
• Stabilizing Reinforcement Learning for Honesty Alignment in Language Models on Deductive Reasoning [27.42733470720954]
  We propose a reinforcement learning method that injects ground truth trajectories into rollouts, preventing early training collapse.<n>Our results demonstrate that this method stabilizes learning and significantly improves the overall reasoning performance.
  arXiv  Detail & Related papers  (2025-11-12T11:34:19Z)
• Provable Benefit of Curriculum in Transformer Tree-Reasoning Post-Training [76.12556589212666]
  We show that curriculum post-training avoids the exponential complexity bottleneck.<n>Under outcome-only reward signals, reinforcement learning finetuning achieves high accuracy with sample complexity.<n>We establish guarantees for test-time scaling, where curriculum-aware querying reduces both reward oracle calls and sampling cost from exponential to order.
  arXiv  Detail & Related papers  (2025-11-10T18:29:54Z)
• Consistency Is Not Always Correct: Towards Understanding the Role of Exploration in Post-Training Reasoning [75.79451512757844]
  Foundation models exhibit broad knowledge but limited task-specific reasoning.<n> RLVR and inference scaling motivate post-training strategies such as RLVR and inference scaling.<n>We show that RLVR induces a squeezing effect, reducing reasoning entropy and forgetting some correct paths.
  arXiv  Detail & Related papers  (2025-11-10T18:25:26Z)
• Unifying Deductive and Abductive Reasoning in Knowledge Graphs with Masked Diffusion Model [64.31242163019242]
  Deductive and abductive reasoning are critical paradigms for analyzing knowledge graphs.<n>We propose a unified framework for Deductive and Abductive Reasoning in Knowledge graphs, called DARK.<n>We show that DARK achieves state-of-the-art performance on both deductive and abductive reasoning tasks.
  arXiv  Detail & Related papers  (2025-10-13T14:34:57Z)
• Controllable Logical Hypothesis Generation for Abductive Reasoning in Knowledge Graphs [54.596180382762036]
  Abductive reasoning in knowledge graphs aims to generate plausible logical hypotheses from observed entities.<n>Due to a lack of controllability, a single observation may yield numerous plausible but redundant or irrelevant hypotheses.<n>We introduce the task of controllable hypothesis generation to improve the practical utility of abductive reasoning.
  arXiv  Detail & Related papers  (2025-05-27T09:36:47Z)
• Failure Modes of LLMs for Causal Reasoning on Narratives [51.19592551510628]
  We investigate the interaction between world knowledge and logical reasoning.<n>We find that state-of-the-art large language models (LLMs) often rely on superficial generalizations.<n>We show that simple reformulations of the task can elicit more robust reasoning behavior.
  arXiv  Detail & Related papers  (2024-10-31T12:48:58Z)

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.