CARE What Fails: Contrastive Anchored-REflection for Verifiable Multimodal Reasoning
Group-relative reinforcement learning with verifiable rewards (RLVR) often wastes the most informative data it already has the failures. When all rollouts are wrong, gradients stall; when one happens to be correct, the update usually ignores why the others are close-but-wrong, and credit can be misassigned to spurious chains. We present CARE (Contrastive Anchored REflection), a failure-centric post-training framework for multimodal reasoning that turns errors into supervision. CARE combines: (i) an anchored-contrastive objective that forms a compact subgroup around the best rollout and a set of semantically proximate hard negatives, performs within-subgroup z-score normalization with negative-only scaling, and includes an all-negative rescue to prevent zero-signal batches; and (ii) Reflection-Guided Resampling (RGR), a one-shot structured self-repair that rewrites a representative failure and re-scores it with the same verifier, converting near-misses into usable positives without any test-time reflection. CARE improves accuracy and training smoothness while explicitly increasing the share of learning signal that comes from failures. On Qwen2.5-VL-7B, CARE lifts macro-averaged accuracy by 4.6 points over GRPO across six verifiable visual-reasoning benchmarks; with Qwen3-VL-8B it reaches competitive or state-of-the-art results on MathVista and MMMU-Pro under an identical evaluation protocol.
Code (0)
등록된 구현이 없습니다.
Tasks
Reinforcement LearningMultimodal ReasoningSimilar Papers 제목 키워드 기반
Contrastive Reflection for Iterative Prompt Optimization
LLM agents are becoming central to information retrieval: they issue retrieval queries, synthesize answers, and increasingly serve as judges for IR evaluation. Improving the prompts that control these agents is an optimi…
Information RetrievalCareMirror: Bringing Caregiver Wellbeing into the Dementia Care Ecosystem
Family caregivers of people living with dementia shoulder emotional and practical responsibilities, yet their own wellbeing often remains peripheral to dementia care. We built CareMirror, an envisioned caregiver wellbein…
MamaBench: Benchmarking LLM Robustness in Maternal and Child Health Diagnosis through Counterfactual Clinical Perturbation
Large language models achieve strong scores on medical benchmarks, yet these benchmarks evaluate each question in isolation, providing no measure of whether a system can distinguish clinically similar presentations requi…
A Doeblin-Anchored Contrastive Chart for Learning Markov Transition Kernels
Learning a Markov transition model is not merely conditional density estimation: the learned object must be a valid transition kernel before it is iterated in downstream dynamics. This paper introduces a Doeblin-anchored…
Density EstimationFrom Narrow to Panoramic Vision: Attention-Guided Cold-Start Reshapes Multimodal Reasoning
The cold-start initialization stage plays a pivotal role in training Multimodal Large Reasoning Models (MLRMs), yet its mechanisms remain insufficiently understood. To analyze this stage, we introduce the Visual Attentio…
Multimodal Reasoning