arXiv AI
The Evidence Anchoring Boundary: Where Visual Interventions Matter and When They Help in Multimodal LLMs
Read the original on arXiv AI →The Flow has not summarised this story yet — read it at arXiv AI.
Related stories
arXiv Computer Vision
Sep 22
Look Where It Counts: A Free, Label-Free Visual Evidence Signal for Fine-Grained Vision-Language Reasoning
The paper introduces a free, label‑free visual evidence signal that improves fine‑grained vision‑language reasoning. By selecting image crops that maximize the model’s answer distribution peak, the method locates answer‑bearing regions without training or annotations, boosting accuracy from 70 % to 85 %. The evidence gap also complements model confidence, enabling better correctness prediction and error flagging.
arXiv AI
Oct 1
VisualNeedle: Benchmarking Active Visual Search in Information-Dense Scenes
arXiv:2605.26380v2 Announce Type: replace-cross Abstract: Frontier multimodal large language models (MLLMs) have been reported to achieve over 90\% accuracy on fine-grained perception benchmarks. How...
Hugging Face Trending Papers
Aug 10
Perception Before Supervision: Self-Contained Visual Distillation from Counterfactual Blind Spots
Self-improvement for multimodal large language models (MLLMs) is typically driven by reward-based methods that provide only coarse scalar feedback. Distillation offers a richer alternative through dense token-level supervision, but in the visual domain it usually depends on privileged context constructed using external annotations and tools, or stronger models.
Hugging Face Trending Papers
Sep 29
Devils in Question Relay: Source-Conditioned Relay Steering to Mitigate Hallucinations in Audio-visual Large Language Models
Audio-visual large language models (AVLLMs) have made remarkable progress in multimodal understanding and reasoning through interactions among visual, auditory, and linguistic information. However, re...
arXiv Computer Vision
Sep 16
SAVOR: Self-Aware Visual Grounding via Confidence-Calibrated Reinforcement Learning for Multimodal Hallucination Mitigation
SAVOR is a training framework for multimodal large language models that adds token and answer confidence to the output schema, optimises a Group Relative Policy Optimisation objective to penalise calibration error and poor abstention, and uses the learned confidence at inference to revisit visual evidence only when uncertain. Experiments on POPE, HallusionBench, AMBER, and MMHal-Bench with InternVL3-8B and Qwen3-VL-8B backbones show that SAVOR reduces hallucination while maintaining general capability on MME and MMBench, achieving lower Expected Calibration Error than DPO and decoding baselines.
arXiv Computer Vision
Sep 3
Who Drives the Probability Game of VLMs? A Temporal Causal Drive Evaluation Framework
The paper introduces a causal and temporal evaluation framework for vision‑language models (VLMs) that tracks how visual input, question text, and generated prefixes influence autoregressive decoding. It defines three step‑indexed causal‑drive metrics—Visual Causal Drive (VCD), Question Causal Drive (QCD), and Prefix Causal Drive (PCD)—using a Structural Causal Model and interventions. Experiments on Qwen3‑VL‑8B‑Instruct and other datasets show a shift from early question and visual guidance to increased reliance on generated prefixes, and demonstrate that QCD and PCD reduce recovery error and improve bias detection.