LLMs lack temporal awareness of medical knowledge, showing gradual performance decline on up-to-date facts, much lower accuracy on historical knowledge (25-54% relative), and inconsistent year-to-year predictions.
hub Mixed citations
Kimi-VL Technical Report
Mixed citation behavior. Most common role is background (64%).
90
Pith papers citing it
Background
64% of classified citations
abstract
We present Kimi-VL, an efficient open-source Mixture-of-Experts (MoE) vision-language model (VLM) that offers advanced multimodal reasoning, long-context understanding, and strong agent capabilities - all while activating only 2.8B parameters in its language decoder (Kimi-VL-A3B). Kimi-VL demonstrates strong performance across challenging domains: as a general-purpose VLM, Kimi-VL excels in multi-turn agent tasks (e.g., OSWorld), matching flagship models. Furthermore, it exhibits remarkable capabilities across diverse challenging vision language tasks, including college-level image and video comprehension, OCR, mathematical reasoning, and multi-image understanding. In comparative evaluations, it effectively competes with cutting-edge efficient VLMs such as GPT-4o-mini, Qwen2.5-VL-7B, and Gemma-3-12B-IT, while surpassing GPT-4o in several key domains. Kimi-VL also advances in processing long contexts and perceiving clearly. With a 128K extended context window, Kimi-VL can process diverse long inputs, achieving impressive scores of 64.5 on LongVideoBench and 35.1 on MMLongBench-Doc. Its native-resolution vision encoder, MoonViT, further allows it to see and understand ultra-high-resolution visual inputs, achieving 83.2 on InfoVQA and 34.5 on ScreenSpot-Pro, while maintaining lower computational cost for common tasks. Building upon Kimi-VL, we introduce an advanced long-thinking variant: Kimi-VL-Thinking-2506. Developed through long chain-of-thought (CoT) supervised fine-tuning (SFT) and reinforcement learning (RL), the latest model exhibits strong long-horizon reasoning capabilities (64.0 on MMMU, 46.3 on MMMU-Pro, 56.9 on MathVision, 80.1 on MathVista, 65.2 on VideoMMMU) while obtaining robust general abilities. Code and models are publicly accessible at https://github.com/MoonshotAI/Kimi-VL.
hub tools
citation-role summary
background 24
baseline 8
method 3
other 1
citation-polarity summary
claims ledger
- abstract We present Kimi-VL, an efficient open-source Mixture-of-Experts (MoE) vision-language model (VLM) that offers advanced multimodal reasoning, long-context understanding, and strong agent capabilities - all while activating only 2.8B parameters in its language decoder (Kimi-VL-A3B). Kimi-VL demonstrates strong performance across challenging domains: as a general-purpose VLM, Kimi-VL excels in multi-turn agent tasks (e.g., OSWorld), matching flagship models. Furthermore, it exhibits remarkable capabilities across diverse challenging vision language tasks, including college-level image and video c
co-cited works
representative citing papers
SenseBench is the first physics-based benchmark with 10K+ instances and dual protocols to evaluate VLMs on remote sensing low-level perception and diagnostic description, revealing domain bias and specific failure modes.
TraceAV-Bench is the first benchmark for multi-hop trajectory reasoning over long audio-visual videos, showing top models reach only 51-68% accuracy with substantial room for improvement.
HM-Bench is the first benchmark for MLLMs on hyperspectral images, showing models struggle with complex spatial-spectral reasoning and perform better with visual PCA images than textual reports.
DriveSpatial benchmark shows the best of 15 VLMs trails humans by 28.4 points on spatiotemporal driving tasks, with cognitive scene construction as the main failure mode.
VideoOdyssey is a new benchmark featuring ultra-long videos (avg. 109 min) across 11 domains with multi-level continuous certificates (avg. 16 min for visual, 12.8 min for audio-visual) to diagnose MLLM limitations in continuous reasoning and omni-modal perception.
SpaceDG introduces the first large-scale degradation-aware spatial reasoning dataset using 3D Gaussian Splatting synthesis, showing that visual degradations impair MLLM performance but finetuning on the data improves robustness and can exceed human levels under degradation.
Proposes Spatial Narrative Score (SNS) evaluation for VLMs' camera motion understanding and introduces CaMo model achieving consistent performance on SNS and direct QA.
PAGER achieves 4.1x higher task success in point-precise geometric GUI control by combining topology-aware planning with precision-aligned reinforcement learning on the new PAGE Bench dataset of 4,906 problems.
ReVision reduces visual token usage by 46% on average in agent trajectories via a learned patch selector and improves success rates by 3% on three benchmarks, showing that history saturation stems from inefficient representations rather than lack of utility.
Multi-grained counting is introduced with five granularity levels, supported by the new KubriCount dataset generated via 3D synthesis and editing, and HieraCount model that combines text and visual exemplars for improved accuracy.
RAPO uses an information-theoretic lower bound on visual gain to select high-entropy reflection anchors and optimizes a chain-masked KL surrogate, delivering gains over baselines on reasoning benchmarks across LVLM backbones.
VIGIL decouples world-state completion from terminal commitment in embodied agents, exposing up to 19.7 pp gaps in benchmark success despite comparable execution across 20 models.
LLM agents reach only 50.6% accuracy on chemical cost estimation within 25% error even with tools, dropping with noise due to parsing, pack selection, and tool-use failures.
Pest-Thinker is a reinforcement learning framework that improves MLLMs' expert-level reasoning on pest morphology via synthesized CoT trajectories, GRPO optimization, and an LLM-judged feature reward on new benchmarks QFSD and AgriInsect.
RouteHijack is a routing-aware jailbreak that identifies safety-critical experts via activation contrast and optimizes suffixes to suppress them, reaching 69.3% average attack success rate on seven MoE LLMs with strong transfer to variants and VLMs.
Introduces QCalEval benchmark showing best zero-shot VLM score of 72.3 on quantum calibration plots, with fine-tuning and in-context learning effects varying by model type.
FCMBench-Video is a new benchmark with 1,200 videos and 11k QA instances for evaluating Video-MLLMs on document video understanding across 28 document types.
Current MLLMs show weak performance on small object understanding tasks, but fine-tuning with the new SOU-Train dataset measurably improves their capabilities.
Visual token pruning in MLLMs fails on complex reasoning due to Relevant Visual Information Shift during decoding, but the DSTP framework fixes it training-free across models.
FeDPM learns and aligns local discrete prototypical memories across domains to create a unified discrete latent space for LLM-based time series foundation models in a federated setting.
Backward token warping in ViT-based MLLMs enables reliable reasoning from nearby viewpoints by preserving semantic coherence better than pixel-wise warping or fine-tuning baselines.
SceneBench shows VLMs lose accuracy on scene-level questions in long videos due to forgetting, and Scene-RAG retrieval improves performance by 2.5%.
ChartNet is a million-scale multimodal dataset for chart understanding created via code-guided synthesis spanning 24 chart types with five aligned modalities per sample.
citing papers explorer
-
Large Language Models Lack Temporal Awareness of Medical Knowledge
LLMs lack temporal awareness of medical knowledge, showing gradual performance decline on up-to-date facts, much lower accuracy on historical knowledge (25-54% relative), and inconsistent year-to-year predictions.
-
SenseBench: A Benchmark for Remote Sensing Low-Level Visual Perception and Description in Large Vision-Language Models
SenseBench is the first physics-based benchmark with 10K+ instances and dual protocols to evaluate VLMs on remote sensing low-level perception and diagnostic description, revealing domain bias and specific failure modes.
-
TraceAV-Bench: Benchmarking Multi-Hop Trajectory Reasoning over Long Audio-Visual Videos
TraceAV-Bench is the first benchmark for multi-hop trajectory reasoning over long audio-visual videos, showing top models reach only 51-68% accuracy with substantial room for improvement.
-
HM-Bench: A Comprehensive Benchmark for Multimodal Large Language Models in Hyperspectral Remote Sensing
HM-Bench is the first benchmark for MLLMs on hyperspectral images, showing models struggle with complex spatial-spectral reasoning and perform better with visual PCA images than textual reports.
-
DRIVESPATIAL: A Benchmark for Spatiotemporal Intelligence in VLMs for Autonomous Driving
DriveSpatial benchmark shows the best of 15 VLMs trails humans by 28.4 points on spatiotemporal driving tasks, with cognitive scene construction as the main failure mode.
-
VideoOdyssey: A Benchmark for Ultra-Long-Context and Omni-Modal Video Understanding
VideoOdyssey is a new benchmark featuring ultra-long videos (avg. 109 min) across 11 domains with multi-level continuous certificates (avg. 16 min for visual, 12.8 min for audio-visual) to diagnose MLLM limitations in continuous reasoning and omni-modal perception.
-
SpaceDG: Benchmarking Spatial Intelligence under Visual Degradation
SpaceDG introduces the first large-scale degradation-aware spatial reasoning dataset using 3D Gaussian Splatting synthesis, showing that visual degradations impair MLLM performance but finetuning on the data improves robustness and can exceed human levels under degradation.
-
CaMo: Camera Motion Grounded Evaluation and Training for Vision-Language Models
Proposes Spatial Narrative Score (SNS) evaluation for VLMs' camera motion understanding and introduces CaMo model achieving consistent performance on SNS and direct QA.
-
PAGER: Bridging the Semantic-Execution Gap in Point-Precise Geometric GUI Control
PAGER achieves 4.1x higher task success in point-precise geometric GUI control by combining topology-aware planning with precision-aligned reinforcement learning on the new PAGE Bench dataset of 4,906 problems.
-
ReVision: Scaling Computer-Use Agents via Temporal Visual Redundancy Reduction
ReVision reduces visual token usage by 46% on average in agent trajectories via a learned patch selector and improves success rates by 3% on three benchmarks, showing that history saturation stems from inefficient representations rather than lack of utility.
-
Count Anything at Any Granularity
Multi-grained counting is introduced with five granularity levels, supported by the new KubriCount dataset generated via 3D synthesis and editing, and HieraCount model that combines text and visual exemplars for improved accuracy.
-
Reflection Anchors for Propagation-Aware Visual Retention in Long-Chain Multimodal Reasoning
RAPO uses an information-theoretic lower bound on visual gain to select high-entropy reflection anchors and optimizes a chain-masked KL surrogate, delivering gains over baselines on reasoning benchmarks across LVLM backbones.
-
Done, But Not Sure: Disentangling World Completion from Self-Termination in Embodied Agents
VIGIL decouples world-state completion from terminal commitment in embodied agents, exposing up to 19.7 pp gaps in benchmark success despite comparable execution across 20 models.
-
Can Agents Price a Reaction? Evaluating LLMs on Chemical Cost Reasoning
LLM agents reach only 50.6% accuracy on chemical cost estimation within 25% error even with tools, dropping with noise due to parsing, pack selection, and tool-use failures.
-
Pest-Thinker: Learning to Think and Reason like Entomologists via Reinforcement Learning
Pest-Thinker is a reinforcement learning framework that improves MLLMs' expert-level reasoning on pest morphology via synthesized CoT trajectories, GRPO optimization, and an LLM-judged feature reward on new benchmarks QFSD and AgriInsect.
-
RouteHijack: Routing-Aware Attack on Mixture-of-Experts LLMs
RouteHijack is a routing-aware jailbreak that identifies safety-critical experts via activation contrast and optimizes suffixes to suppress them, reaching 69.3% average attack success rate on seven MoE LLMs with strong transfer to variants and VLMs.
-
QCalEval: Benchmarking Vision-Language Models for Quantum Calibration Plot Understanding
Introduces QCalEval benchmark showing best zero-shot VLM score of 72.3 on quantum calibration plots, with fine-tuning and in-context learning effects varying by model type.
-
FCMBench-Video: Benchmarking Document Video Intelligence
FCMBench-Video is a new benchmark with 1,200 videos and 11k QA instances for evaluating Video-MLLMs on document video understanding across 28 document types.
-
Can Multimodal Large Language Models Truly Understand Small Objects?
Current MLLMs show weak performance on small object understanding tasks, but fine-tuning with the new SOU-Train dataset measurably improves their capabilities.
-
Why and When Visual Token Pruning Fails? A Study on Relevant Visual Information Shift in MLLMs Decoding
Visual token pruning in MLLMs fails on complex reasoning due to Relevant Visual Information Shift during decoding, but the DSTP framework fixes it training-free across models.
-
Discrete Prototypical Memories for Federated Time Series Foundation Models
FeDPM learns and aligns local discrete prototypical memories across domains to create a unified discrete latent space for LLM-based time series foundation models in a federated setting.
-
Token Warping Helps MLLMs Look from Nearby Viewpoints
Backward token warping in ViT-based MLLMs enables reliable reasoning from nearby viewpoints by preserving semantic coherence better than pixel-wise warping or fine-tuning baselines.
-
Seeing the Scene Matters: Revealing Forgetting in Video Understanding Models with a Scene-Aware Long-Video Benchmark
SceneBench shows VLMs lose accuracy on scene-level questions in long videos due to forgetting, and Scene-RAG retrieval improves performance by 2.5%.
-
ChartNet: A Million-Scale, High-Quality Multimodal Dataset for Robust Chart Understanding
ChartNet is a million-scale multimodal dataset for chart understanding created via code-guided synthesis spanning 24 chart types with five aligned modalities per sample.
-
VISTA-Bench: Do Vision-Language Models Really Understand Visualized Text as Well as Pure Text?
VISTA-Bench shows vision-language models degrade on visualized text in images compared to equivalent pure text, with larger gaps under increased perceptual difficulty.
-
Weather-R1: Logically Consistent Reinforcement Fine-Tuning for Multimodal Reasoning in Meteorology
Weather-R1 is a multimodal reasoning model for meteorology that uses logical consistency rewards during reinforcement fine-tuning to cut self-contradictory outputs and raises benchmark accuracy by 9.8 points over baselines.
-
ProcObject-10K: Benchmarking Object-Centric Procedural Understanding in Instructional Videos
ProcObject-10K is the first benchmark for object-centric procedural reasoning in videos that exposes a large gap where models answer questions plausibly but fail to ground their answers in the correct video segments.
-
From Charts to Code: A Hierarchical Benchmark for Multimodal Models
Chart2Code is a hierarchical benchmark with reproduction, editing, and table-to-chart tasks across 22 chart types that shows even top models like GPT-5 achieve low scores of 0.57 on code evaluation and 0.22 on chart quality for editing tasks.
-
High-Resolution Visual Reasoning via Multi-Turn Grounding-Based Reinforcement Learning
MGPO elicits grounding in LMMs via multi-turn RL with binary rewards, yielding 5.4% and 5.2% gains on MME-Realworld and V* Bench and surpassing GPT-4o on the latter after training on 21K samples.
-
MMSearch-R1: Incentivizing LMMs to Search
MMSearch-R1 uses reinforcement learning to train multimodal models for on-demand multi-turn internet search with image and text tools, outperforming same-size RAG baselines and matching larger ones while cutting search calls by over 30%.
-
PuzzleWorld: A Benchmark for Multimodal, Open-Ended Reasoning in Puzzlehunts
PuzzleWorld benchmark reveals state-of-the-art AI models solve only 18% of complex puzzlehunt problems with 40% stepwise accuracy, matching novices but trailing enthusiasts, while fine-tuning on traces yields modest gains.
-
Video-Holmes: Can MLLM Think Like Holmes for Complex Video Reasoning?
Video-Holmes benchmark shows top MLLMs achieve at most 45% accuracy on tasks needing integration of multiple clues from suspense films, unlike existing perception-focused tests.
-
SpatialScore: Towards Comprehensive Evaluation for Spatial Intelligence
Presents SpatialScore benchmark for MLLM spatial reasoning, evaluates 49 models showing large human gap, and supplies SpatialCorpus plus SpatialAgent to improve performance.
-
SpaceR: Reinforcing MLLMs in Video Spatial Reasoning
SpaceR uses a new verifiable dataset and map-imagination-augmented RLVR to reach SOTA spatial reasoning accuracy in MLLMs, exceeding GPT-4o on VSI-Bench.
-
OCRBench v2: An Improved Benchmark for Evaluating Large Multimodal Models on Visual Text Localization and Reasoning
OCRBench v2 is a new benchmark with four times more tasks than prior versions that reveals most large multimodal models score below 50 out of 100 on visual text tasks and share five specific weaknesses.
-
JUDO: A Juxtaposed Domain-Oriented Multimodal Reasoner for Industrial Anomaly QA
JUDO enhances large multimodal models for industrial anomaly QA by juxtaposing query images with normal ones for visual comparison and using SFT plus GRPO with tailored rewards to inject domain knowledge, outperforming Qwen2.5-VL-7B and GPT-4o on the MMAD benchmark.
-
Attention Hijacking: Response Manipulation Across Queries in Vision-Language Models
Attention Hijacking is a new attack that improves cross-query transferability in VLMs by explicitly steering internal attention to a persistent image-dominant pattern.
-
When Efficiency Backfires: Cascading LLMs Trigger Cascade Failure under Adversarial Attack
LLM cascade systems are vulnerable to a new adversarial attack that simultaneously degrades accuracy and destroys the intended cost savings by targeting both the lightweight models and the escalation decision mechanism.
-
Artificial Intolerance: Stigmatizing Language in Clinical Documentation Skews Large Language Model Decision-Making
Frontier LLMs exhibit bias from stigmatizing language in clinical vignettes across four conditions, skewing decisions toward less aggressive management, with limited mitigation from Chain-of-Thought or self-debiasing prompts.
-
Attribute-Grounded Selective Reasoning for Artwork Emotion Understanding with Multimodal Large Language Models
Proposes AGSR and the FAB-G supervised multi-agent framework that predicts attribute salience from human annotations to constrain MLLM emotion reasoning, yielding gains on EmoArt and cross-dataset tests.
-
UAM: A Dual-Stream Perspective on Forgetting in VLA Training
UAM adds a Dorsal Expert initialized from a generative model and trained on visual dynamics prediction to preserve over 95% of VLM multimodal ability in VLA training while achieving top success rates on manipulation tasks including OOD cases.
-
Learning to See What You Need: Gaze Attention for Multimodal Large Language Models
Gaze Attention groups visual embeddings into selectable regions and dynamically restricts attention to task-relevant ones, matching dense baselines with up to 90% fewer visual KV entries via added context tokens.
-
NanoResearch: Co-Evolving Skills, Memory, and Policy for Personalized Research Automation
NanoResearch introduces a tri-level co-evolving framework of skills, memory, and policy to personalize LLM-powered research automation across projects and users.
-
SpaceMind++: Toward Allocentric Cognitive Maps for Spatially Grounded Video MLLMs
SpaceMind++ adds an explicit voxelized allocentric cognitive map and coordinate-guided fusion to video MLLMs, claiming SOTA on VSI-Bench and improved out-of-distribution generalization on three other 3D benchmarks.
-
Dimension-Free Saddle-Point Escape in Muon
Muon achieves dimension-free saddle-point escape through non-linear spectral shaping, resolvent calculus, and structural incoherence, yielding an algebraically dimension-free escape bound.
-
Beyond Thinking: Imagining in 360$^\circ$ for Humanoid Visual Search
Imagining in 360° decouples visual search into a single-step probabilistic semantic layout predictor and an actor, removing the need for multi-turn CoT reasoning and trajectory annotations while improving efficiency in 360° environments.
-
LLaVA-UHD v4: What Makes Efficient Visual Encoding in MLLMs?
LLaVA-UHD v4 reduces visual-encoding FLOPs by 55.8% for high-resolution images in MLLMs via slice-based encoding plus intra-ViT early compression while matching or exceeding baseline performance on document, OCR, and VQA benchmarks.
-
LiteGUI: Distilling Compact GUI Agents with Reinforcement Learning
LiteGUI trains 2B/3B-scale GUI agents via SFT-free guided on-policy distillation and multi-solution dual-level GRPO to reach SOTA lightweight performance and compete with larger models.
-
Adaptive Inverted-Index Routing for Granular Mixtures-of-Experts
AIR-MoE introduces a two-stage inverted-index routing method based on vector quantization that approximates optimal expert selection for granular MoE models at lower cost and with empirical performance gains.
-
Persistent Visual Memory: Sustaining Perception for Deep Generation in LVLMs
PVM adds a parallel branch to LVLMs that directly supplies visual embeddings to prevent attention decay over long generated sequences, yielding accuracy gains on reasoning tasks with minimal overhead.