Perception Without Engagement: Dissecting the Causal Discovery Deficit in LMMs

• URL: http://arxiv.org/abs/2605.09422v1
• Date: Sun, 10 May 2026 08:48:58 GMT
• Title: Perception Without Engagement: Dissecting the Causal Discovery Deficit in LMMs
• Abstract summary: We introduce ProCauEval, a perturbation-based evaluation protocol that shifts from outcome assessment to mechanism diagnosis.<n>We find that models faithfully perceive video content yet systematically underexploit it during causal reasoning.<n>We propose Anti-Distillation Policy Optimization (ADPO), a reinforcement learning framework built on negative teacher alignment.
• Score: 68.58207076756237
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Although Large Multimodal Models (LMMs) have achieved strong performance on general video understanding, their susceptibility to textual prior shortcuts during causal discovery has been recognized as a critical deficit. The underlying mechanisms of this phenomenon remain incompletely understood, as existing benchmarks only measure response accuracy without revealing the sources and extent of the deficit. We introduce ProCauEval, a perturbation-based evaluation protocol that shifts from outcome assessment to mechanism diagnosis, probing causal discovery through five controlled configurations that systematically manipulate visual and textual modalities to decompose their respective contributions to model behavior and dissect the failure modes. Evaluating 17 mainstream LMMs, we find that models faithfully perceive video content yet systematically underexploit it during causal reasoning. We further observe that stronger post-training amplifies rather than mitigates textual prior reliance, and that higher baseline performance correlates with greater fragility under perturbation. To address these, we propose Anti-Distillation Policy Optimization (ADPO), a reinforcement learning framework built on negative teacher alignment, which augments GRPO by explicitly pushing the policy away from a prior-only counterfactual teacher induced by visual corruption. Specifically, ADPO maximizes the divergence between the policy distributions conditioned on the original and visually corrupted inputs, thereby forcing the model to ground its reasoning in visual evidence rather than textual shortcuts. Extensive experiments show that ADPO improves visual engagement without sacrificing fundamental comprehension, thus offering a preliminary step toward reliable causal discovery.

Related papers

• Evidence-Grounded Trustworthy Multimodal Reasoning and Evaluation Benchmark in Complex Urban Scenes [37.743333035442426]
  This paper introduces AD2-Bench, which decomposes reasoning into a structured Chain of Evidence (CoE)<n>Building on this perspective, the authors identify two primary causes of reasoning failure: Spatial Ambiguity and Semantic Uncertainty.<n>They then propose Evidence-grounded Visual Reasoning (EGVOR), which replaces implicit reasoning with the explicit generation of Evidence Atoms.
  arXiv  Detail & Related papers  (2026-08-11T14:23:07Z)
• CFPO: Counterfactual Policy Optimization for Multimodal Reasoning [6.719961045766247]
  We propose CounterFactual Policy Optimization (CFPO), a framework that enforces causal consistency between visual perception and textual reasoning.<n>CFPO integrates with standard algorithms like GRPO and DAPO without requiring external reward models or additional supervision.<n>It significantly improves reasoning fidelity, achieving consistent gains of 3.17%-6.25% over standard RL baselines and 1.32%-2.13% over the state-of-the-art perception-aware method (PAPO)
  arXiv  Detail & Related papers  (2026-06-22T11:51:51Z)
• The Cost of Context: Mitigating Textual Bias in Multimodal Retrieval-Augmented Generation [18.83395586542378]
  Multimodal Large Language Models (MLLMs) are increasingly integrated with Retrieval-Augmented Generation (RAG)<n>We identify and formalize the phenomenon of recorruption, where the introduction of even perfectly accurate "oracle" context causes a capable model to abandon an initially correct prediction.<n>We propose Bottleneck Attention Intervention for Recovery (BAIR), a parameter-free, inference-time framework that restores visual saliency and applies position-aware penalties to textual distractors.
  arXiv  Detail & Related papers  (2026-05-07T02:27:04Z)
• Enhancing Multimodal In-Context Learning via Inductive-Deductive Reasoning [65.15766304205657]
  In-context learning (ICL) allows large models to adapt to tasks using a few examples, yet its extension to vision-language models (VLMs) remains fragile.<n>Our analysis reveals that the fundamental limitation lies in an inductive gap, models often produce correct answers from flawed reasoning.<n>We introduce a framework that restructures multimodal ICL as a principled inductive-deductive process.
  arXiv  Detail & Related papers  (2026-05-04T09:18:19Z)
• DO-Bench: An Attributable Benchmark for Diagnosing Object Hallucination in Vision-Language Models [17.5535076530203]
  Object level hallucination remains a central reliability challenge for vision language models.<n>We introduce DO-Bench, a controlled diagnostic benchmark that isolates sources through structured interventions.<n>We show that object hallucination reflects heterogeneous, mechanism dependent failure patterns beyond aggregate accuracy.
  arXiv  Detail & Related papers  (2026-04-18T06:54:02Z)
• Adaptive Uncertainty-Aware Tree Search for Robust Reasoning [30.006968507593452]
  Inference-time reasoning has advanced the capabilities of Large Language Models (LLMs) in complex problem-solving.<n>A prevalent approach involves external search guided by Process Reward Models (PRMs)<n>We provide empirical evidence that PRMs exhibit high uncertainty and unreliable scoring on out-of-distribution (OOD) samples.<n>We then establish a theoretical framework proving that while standard search incurs linear regret accumulation, an uncertainty-aware strategy can achieve sublinear regret.
  arXiv  Detail & Related papers  (2026-02-06T08:40:05Z)
• Think Locally, Explain Globally: Graph-Guided LLM Investigations via Local Reasoning and Belief Propagation [5.191980417814362]
  LLM agents excel when environments are mostly static and the needed information fits in a model's context window.<n>ReAct-style agents are especially brittle in this regime.<n>We propose EoG, a framework in which an LLM performs bounded local evidence mining and labeling (cause vs symptom) while a deterministic controller manages, state, and belief propagation to compute a minimal explanatory frontier.
  arXiv  Detail & Related papers  (2026-01-25T17:27:19Z)
• Analyzing Reasoning Consistency in Large Multimodal Models under Cross-Modal Conflicts [74.47786985522762]
  We identify a critical failure mode termed textual inertia, where models tend to blindly adhere to the erroneous text while neglecting conflicting visual evidence.<n>We propose the LogicGraph Perturbation Protocol that structurally injects perturbations into the reasoning chains of diverse LMMs.<n>Results reveal that models successfully self-correct in less than 10% of cases and predominantly succumb to blind textual error propagation.
  arXiv  Detail & Related papers  (2026-01-07T16:39:34Z)
• Learning What to Attend First: Modality-Importance-Guided Reasoning for Reliable Multimodal Emotion Understanding [50.014363382140324]
  Modality Importance (MI) is a simple yet effective mechanism for identifying the emotion-dominant modality.<n>MIGR reorganizes reasoning sequences so that explanations begin from the modality most critical to the target emotion.<n>Results show that MIGR substantially improves reasoning reliability, decreasing instances of correct predictions accompanied by emotionally inconsistent explanations.
  arXiv  Detail & Related papers  (2025-12-02T12:29:41Z)
• VAR: Visual Attention Reasoning via Structured Search and Backtracking [49.427842994857635]
  We introduce Visual Attention Reasoning, a framework that recasts grounded reasoning as a structured search.<n> VAR decomposes the reasoning process into two key stages: traceable evidence grounding and search-based chain-of-thought.<n>We show that our 7B model, VAR-7B, sets a new state-of-the-art on a comprehensive suite of hallucination and safety benchmarks.
  arXiv  Detail & Related papers  (2025-10-21T13:18:44Z)
• Perception-Consistency Multimodal Large Language Models Reasoning via Caption-Regularized Policy Optimization [72.30168853571216]
  multimodal large language models excel at tasks that integrate visual perception with symbolic reasoning.<n>CapPO integrates two key mechanisms: (1) a caption-based consistency regularization, which minimizes the divergence between responses conditioned on raw images and those conditioned on captions, and (2) a KL-weighted advantage estimation scheme, which adaptively scales reinforcement signals to strengthen perceptually consistent trajectories.
  arXiv  Detail & Related papers  (2025-09-26T04:32:26Z)
• Lost at the Beginning of Reasoning [85.17612793300238]
  We show that the first reasoning step exerts a disproportionately large influence on the final prediction.<n>We propose an efficient sampling strategy that leverages a reward model to identify and retain high-quality first reasoning steps.
  arXiv  Detail & Related papers  (2025-06-27T09:53:57Z)

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.