MAPO improves multimodal chain-of-thought reasoning by requiring explicit textual descriptions of visual tool results and using a novel advantage estimator that combines semantic alignment with task rewards.
V?: Guided visual search as a core mechanism in multimodal llms
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 6verdicts
UNVERDICTED 6roles
background 1polarities
background 1representative citing papers
An RL-based questioner agent adaptively generates queries to discover novel failure modes in VLMs without human intervention.
DeepEyesV2 uses a two-stage cold-start plus reinforcement learning pipeline to produce an agentic multimodal model that adaptively invokes tools and outperforms direct RL on real-world reasoning benchmarks.
ERASE prunes 85% of vision tokens in Qwen2.5-VL-7B while retaining 89.46% accuracy, outperforming prior methods that retain only 78.1%.
Mini-o3 scales visual search reasoning to tens of interaction turns via a new probe dataset, iterative trajectory collection, and over-turn masking in RL, claiming SOTA performance while training only up to six turns.
LENS is a new multi-level benchmark dataset for evaluating MLLMs on perception-to-reasoning tasks using the same images across all levels with recent social media content.
citing papers explorer
-
Walk the Talk: Bridging the Reasoning-Action Gap for Thinking with Images via Multimodal Agentic Policy Optimization
MAPO improves multimodal chain-of-thought reasoning by requiring explicit textual descriptions of visual tool results and using a novel advantage estimator that combines semantic alignment with task rewards.
-
Discovering Failure Modes in Vision-Language Models using RL
An RL-based questioner agent adaptively generates queries to discover novel failure modes in VLMs without human intervention.
-
DeepEyesV2: Toward Agentic Multimodal Model
DeepEyesV2 uses a two-stage cold-start plus reinforcement learning pipeline to produce an agentic multimodal model that adaptively invokes tools and outperforms direct RL on real-world reasoning benchmarks.
-
ERASE: Eliminating Redundant Visual Tokens via Adaptive Two-Stage Token Pruning
ERASE prunes 85% of vision tokens in Qwen2.5-VL-7B while retaining 89.46% accuracy, outperforming prior methods that retain only 78.1%.
-
Mini-o3: Scaling Up Reasoning Patterns and Interaction Turns for Visual Search
Mini-o3 scales visual search reasoning to tens of interaction turns via a new probe dataset, iterative trajectory collection, and over-turn masking in RL, claiming SOTA performance while training only up to six turns.
-
LENS: Multi-level Evaluation of Multimodal Reasoning with Large Language Models
LENS is a new multi-level benchmark dataset for evaluating MLLMs on perception-to-reasoning tasks using the same images across all levels with recent social media content.