MuseBench shows state-of-the-art MLLMs achieve only 48.29% accuracy on intent-level audiovisual arts understanding versus 87.18% for human experts.
hub Mixed citations
Videorft: Incentivizing video reasoning capability in mllms via reinforced fine-tuning
Mixed citation behavior. Most common role is background (67%).
22
Pith papers citing it
Background
67% of classified citations
hub tools
citation-role summary
background 5
baseline 1
citation-polarity summary
representative citing papers
VideoKR supplies 315K knowledge-intensive video reasoning examples and a dedicated benchmark, with experiments indicating post-training gains on reasoning tasks that require both video content and external knowledge.
Act2See trains VLMs via supervised fine-tuning on verified reasoning traces to interleave active frame calls within text CoTs, yielding SOTA results on video reasoning benchmarks.
VideoP2R separates perception and reasoning in a process-aware RFT pipeline with a new CoT dataset and PA-GRPO rewards, reaching SOTA on six of seven video benchmarks.
Introduces TennisTV benchmark for evaluating 17 MLLMs on tennis video understanding from stroke-level to rally-level tasks with automated pipelines and human verification.
CineCap combines structured reasoning and RL rewards to outperform baselines on cinematographic video captioning using a new 472-pair benchmark.
CARE uses exponential moving average competence estimates to progressively shift RL rewards from exploration-oriented long reasoning to efficiency-oriented concise reasoning in video-MLLMs, with batch normalization and posterior amplification, yielding accuracy gains and shorter traces.
CF-GRPO creates a consensus frame prior from intrinsic video cues and aligns it with model frame-use scores via a reward signal to enable evidence-aware reasoning in Video-MLLMs without temporal annotations.
DynFrame introduces tokenized learnable span-density retrieval and Segment-Decoupled GRPO in video MLLMs, achieving competitive or SOTA results on six benchmarks with 4B and 8B models.
Introduces the first benchmark for metaphorical video understanding, identifies MLLM weaknesses in cross-domain mapping, and proposes an inference-time enhancement using a knowledge graph.
ParaVT introduces the first multi-agent RL framework for parallel video tool calling in LMMs, using PARA-GRPO to resolve the Tool Prior Paradox and achieve +7.9% average improvement over Qwen3-VL baseline across six benchmarks.
VideoSeeker integrates agentic reasoning and visual prompts into LVLMs via automated data synthesis, cold-start supervision, and RL training, yielding +13.7% gains on instance-level video tasks over baselines including GPT-4o.
RLER trains video-reasoning models with three task-driven RL rewards for evidence production and elects the best answer from a few candidates via evidence consistency scoring, yielding 6.3% average gains on eight benchmarks.
AdaTooler-V trains MLLMs to adaptively use vision tools via AT-GRPO reinforcement learning and new datasets, reaching 89.8% on V* and outperforming GPT-4o.
LongVT adds native video-cropping tool calling to LMMs for interleaved multimodal chain-of-tool-thought reasoning on long videos and releases VideoSIAH data for training and evaluation.
MUSEG applies timestamp-aware multi-segment grounding with a phased-reward RL recipe to boost temporal grounding and time-sensitive video QA performance in MLLMs.
VISD proposes structured self-distillation with a multi-dimensional judge model and direction-magnitude decoupling to improve token-level credit assignment and convergence speed in VideoLLM reasoning training.
TempR1 applies temporal-aware multi-task RL using GRPO and three types of localization rewards to achieve SOTA temporal understanding in MLLMs with synergistic gains from joint optimization.
This is a survey that frames video MLLM research via a human-view formulation of perceptual representations, memory states, reasoning traces, and predictions, then reviews methods, datasets, benchmarks, and open problems.
CurEvo integrates curriculum guidance into self-evolution to structure autonomous improvement of video understanding models, yielding gains on VideoQA benchmarks.
citing papers explorer
-
MuseBench: Benchmarking Intent-Level Audiovisual Arts Understanding in MLLMs
MuseBench shows state-of-the-art MLLMs achieve only 48.29% accuracy on intent-level audiovisual arts understanding versus 87.18% for human experts.
-
VideoKR: Towards Knowledge- and Reasoning-Intensive Video Understanding
VideoKR supplies 315K knowledge-intensive video reasoning examples and a dedicated benchmark, with experiments indicating post-training gains on reasoning tasks that require both video content and external knowledge.
-
Act2See: Emergent Active Visual Perception for Video Reasoning
Act2See trains VLMs via supervised fine-tuning on verified reasoning traces to interleave active frame calls within text CoTs, yielding SOTA results on video reasoning benchmarks.
-
VIDEOP2R: Video Understanding from Perception to Reasoning
VideoP2R separates perception and reasoning in a process-aware RFT pipeline with a new CoT dataset and PA-GRPO rewards, reaching SOTA on six of seven video benchmarks.
-
TennisTV: Do Multimodal Large Language Models Understand Tennis Rallies?
Introduces TennisTV benchmark for evaluating 17 MLLMs on tennis video understanding from stroke-level to rally-level tasks with automated pipelines and human verification.
-
CineCap: Structured Reasoning with Spatio-Temporal Anchors for Cinematographic Video Captioning
CineCap combines structured reasoning and RL rewards to outperform baselines on cinematographic video captioning using a new 472-pair benchmark.
-
CARE: Competence-Aware Reward Shaping for Adaptive Reasoning Length in Video-MLLMs
CARE uses exponential moving average competence estimates to progressively shift RL rewards from exploration-oriented long reasoning to efficiency-oriented concise reasoning in video-MLLMs, with batch normalization and posterior amplification, yielding accuracy gains and shorter traces.
-
Reasoning as Intersection: Consensus-Frame Alignment for Visual Focus in Video-MLLMs
CF-GRPO creates a consensus frame prior from intrinsic video cues and aligns it with model frame-use scores via a reward signal to enable evidence-aware reasoning in Video-MLLMs without temporal annotations.
-
DynFrame: Adaptive Reasoning-Driven Multimodal Framework with Dynamic Frame Augmentation for Complex Video Understanding
DynFrame introduces tokenized learnable span-density retrieval and Segment-Decoupled GRPO in video MLLMs, achieving competitive or SOTA results on six benchmarks with 4B and 8B models.
-
MetaphorVU: Towards Metaphorical Video Understanding
Introduces the first benchmark for metaphorical video understanding, identifies MLLM weaknesses in cross-domain mapping, and proposes an inference-time enhancement using a knowledge graph.
-
ParaVT: Taming the Tool Prior Paradox for Parallel Tool Use in Agentic Video Reinforcement Learning
ParaVT introduces the first multi-agent RL framework for parallel video tool calling in LMMs, using PARA-GRPO to resolve the Tool Prior Paradox and achieve +7.9% average improvement over Qwen3-VL baseline across six benchmarks.
-
VideoSeeker: Incentivizing Instance-level Video Understanding via Native Agentic Tool Invocation
VideoSeeker integrates agentic reasoning and visual prompts into LVLMs via automated data synthesis, cold-start supervision, and RL training, yielding +13.7% gains on instance-level video tasks over baselines including GPT-4o.
-
Reinforce to Learn, Elect to Reason: A Dual Paradigm for Video Reasoning
RLER trains video-reasoning models with three task-driven RL rewards for evidence production and elects the best answer from a few candidates via evidence consistency scoring, yielding 6.3% average gains on eight benchmarks.
-
AdaTooler-V: Adaptive Tool-Use for Images and Videos
AdaTooler-V trains MLLMs to adaptively use vision tools via AT-GRPO reinforcement learning and new datasets, reaching 89.8% on V* and outperforming GPT-4o.
-
LongVT: Incentivizing "Thinking with Long Videos" via Native Tool Calling
LongVT adds native video-cropping tool calling to LMMs for interleaved multimodal chain-of-tool-thought reasoning on long videos and releases VideoSIAH data for training and evaluation.
-
MUSEG: Reinforcing Video Temporal Understanding via Timestamp-Aware Multi-Segment Grounding
MUSEG applies timestamp-aware multi-segment grounding with a phased-reward RL recipe to boost temporal grounding and time-sensitive video QA performance in MLLMs.
-
VISD: Enhancing Video Reasoning via Structured Self-Distillation
VISD proposes structured self-distillation with a multi-dimensional judge model and direction-magnitude decoupling to improve token-level credit assignment and convergence speed in VideoLLM reasoning training.
-
TempR1: Improving Temporal Understanding of MLLMs via Temporal-Aware Multi-Task Reinforcement Learning
TempR1 applies temporal-aware multi-task RL using GRPO and three types of localization rewards to achieve SOTA temporal understanding in MLLMs with synergistic gains from joint optimization.
-
Watch, Remember, Reason: Human-View Video Understanding with MLLMs
This is a survey that frames video MLLM research via a human-view formulation of perceptual representations, memory states, reasoning traces, and predictions, then reviews methods, datasets, benchmarks, and open problems.
-
CurEvo: Curriculum-Guided Self-Evolution for Video Understanding
CurEvo integrates curriculum guidance into self-evolution to structure autonomous improvement of video understanding models, yielding gains on VideoQA benchmarks.
- Tracking the Truth: Object-Centric Spatio-Temporal Monitoring for Video Large Language Models
- Auditing Automated Evaluation, Error Propagation, and Runtime Mitigation in Tool-Using Language Agents