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.
super hub Baseline reference
MiniCPM-V: A GPT-4V Level MLLM on Your Phone
Baseline reference. 62% of citing Pith papers use this work as a benchmark or comparison.
104
Pith papers citing it
Baseline
62% of classified citations
abstract
The recent surge of Multimodal Large Language Models (MLLMs) has fundamentally reshaped the landscape of AI research and industry, shedding light on a promising path toward the next AI milestone. However, significant challenges remain preventing MLLMs from being practical in real-world applications. The most notable challenge comes from the huge cost of running an MLLM with a massive number of parameters and extensive computation. As a result, most MLLMs need to be deployed on high-performing cloud servers, which greatly limits their application scopes such as mobile, offline, energy-sensitive, and privacy-protective scenarios. In this work, we present MiniCPM-V, a series of efficient MLLMs deployable on end-side devices. By integrating the latest MLLM techniques in architecture, pretraining and alignment, the latest MiniCPM-Llama3-V 2.5 has several notable features: (1) Strong performance, outperforming GPT-4V-1106, Gemini Pro and Claude 3 on OpenCompass, a comprehensive evaluation over 11 popular benchmarks, (2) strong OCR capability and 1.8M pixel high-resolution image perception at any aspect ratio, (3) trustworthy behavior with low hallucination rates, (4) multilingual support for 30+ languages, and (5) efficient deployment on mobile phones. More importantly, MiniCPM-V can be viewed as a representative example of a promising trend: The model sizes for achieving usable (e.g., GPT-4V) level performance are rapidly decreasing, along with the fast growth of end-side computation capacity. This jointly shows that GPT-4V level MLLMs deployed on end devices are becoming increasingly possible, unlocking a wider spectrum of real-world AI applications in the near future.
hub tools
citation-role summary
baseline 16
background 10
dataset 2
method 1
citation-polarity summary
claims ledger
- abstract The recent surge of Multimodal Large Language Models (MLLMs) has fundamentally reshaped the landscape of AI research and industry, shedding light on a promising path toward the next AI milestone. However, significant challenges remain preventing MLLMs from being practical in real-world applications. The most notable challenge comes from the huge cost of running an MLLM with a massive number of parameters and extensive computation. As a result, most MLLMs need to be deployed on high-performing cloud servers, which greatly limits their application scopes such as mobile, offline, energy-sensitive
authors
co-cited works
representative citing papers
MedHorizon benchmark reveals current multimodal LLMs achieve only 41.1% accuracy on long medical videos due to failures in sparse evidence retrieval and procedural reasoning.
SpikeMLLM is the first spike-based MLLM framework that maintains near-lossless performance under aggressive timestep compression and delivers 9x throughput and 25x power efficiency gains via a custom RTL accelerator.
EgoSound is a new benchmark with 7315 QA pairs across seven tasks to evaluate egocentric sound understanding in multimodal large language models.
ErrorRadar is a new benchmark of 2,500 multimodal K-12 math problems for MLLM error step identification and categorization, where GPT-4o trails human experts by ~10%.
Molmo VLMs trained on newly collected PixMo open datasets achieve state-of-the-art performance among open-weight models and surpass multiple proprietary VLMs including Claude 3.5 Sonnet and Gemini 1.5 Pro.
MMMU-Pro is a stricter multimodal benchmark that removes text-only solvable questions, augments options, and requires reading text from images, yielding substantially lower model scores of 16.8-26.9%.
Introduces the Grounded Personality Reasoning task and MM-OCEAN dataset to show that MLLMs frequently produce correct Big Five personality ratings without grounding them in observable video evidence.
LatentOmni proposes a latent-space cross-modal reasoning framework that uses feature-level supervision and Omni-Sync Position Embedding to align and synchronize audio-visual latents, supported by a new 35K interleaved reasoning dataset and showing gains over text CoT baselines.
CBT-Audio dataset shows that adding audio input improves distress intensity estimation over transcripts alone for 8 of 10 audio language models, with clearest gains when verbal content and vocal delivery diverge.
Omni-DuplexEval creates a new benchmark and LLM-as-a-Judge framework for real-time duplex omni-modal interaction, revealing that current models score below 40% overall and struggle especially with proactive responses.
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.
A proposer-solver agent pair achieves supervised-level video temporal grounding and fine-grained captioning from 2.5K unlabeled videos via self-reinforcing evolution.
TOC-Bench is a new diagnostic benchmark that reveals major weaknesses in temporal object consistency for Video-LLMs, including event counting, ordering, identity reasoning, and hallucination avoidance.
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.
CGC improves fine-grained multi-image understanding in MLLMs by constructing contrastive training instances from existing single-image annotations and adding a rule-based spatial reward, achieving SOTA on MIG-Bench and VLM2-Bench with transfer gains to other multimodal tasks.
SportsTime benchmark and CoTR method improve multimodal AI's temporal compositional reasoning and evidence grounding in long-form sports videos.
A new benchmark converts video clips into shared grounded event records and tests models across physics, semantic, and control prompts under original, shuffled, ablated, and masked conditions, finding selective robustness and weak spatial performance.
WildFireVQA is a new large-scale visual question answering benchmark that pairs RGB imagery with radiometric thermal measurements for aerial wildfire monitoring across six task categories.
Ghost-100 benchmark shows prompt tone drives hallucination rates and intensities in VLMs, with non-monotonic peaks at intermediate pressure and task-specific differences that aggregate metrics hide.
Introduces culture-aware humorous captioning task and staged alignment framework that improves contextual fit and balances image relevance with humor in multimodal LLMs.
OASIS organizes streaming video into hierarchical events and retrieves memory on-demand via intent-driven refinement to improve long-horizon accuracy and compositional reasoning with bounded token costs.
MNAFT identifies language-agnostic and language-specific neurons via activation analysis and selectively fine-tunes only relevant ones in MLLMs to close the modality gap and outperform full fine-tuning and other methods on image translation benchmarks.
Paza is a zero-shot, model-agnostic pipeline that uses behavioral pre-filters on cheap object and pose models to trigger expensive VLMs only when needed, delivering 89.5% precision and 92.8% specificity on a synthesized shoplifting dataset at far lower cost than trained alternatives.
citing papers explorer
-
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.
-
MedHorizon: Towards Long-context Medical Video Understanding in the Wild
MedHorizon benchmark reveals current multimodal LLMs achieve only 41.1% accuracy on long medical videos due to failures in sparse evidence retrieval and procedural reasoning.
-
SpikeMLLM: Spike-based Multimodal Large Language Models via Modality-Specific Temporal Scales and Temporal Compression
SpikeMLLM is the first spike-based MLLM framework that maintains near-lossless performance under aggressive timestep compression and delivers 9x throughput and 25x power efficiency gains via a custom RTL accelerator.
-
EgoSound: Benchmarking Sound Understanding in Egocentric Videos
EgoSound is a new benchmark with 7315 QA pairs across seven tasks to evaluate egocentric sound understanding in multimodal large language models.
-
ErrorRadar: Benchmarking Complex Mathematical Reasoning of Multimodal Large Language Models Via Error Detection
ErrorRadar is a new benchmark of 2,500 multimodal K-12 math problems for MLLM error step identification and categorization, where GPT-4o trails human experts by ~10%.
-
Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Models
Molmo VLMs trained on newly collected PixMo open datasets achieve state-of-the-art performance among open-weight models and surpass multiple proprietary VLMs including Claude 3.5 Sonnet and Gemini 1.5 Pro.
-
MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmark
MMMU-Pro is a stricter multimodal benchmark that removes text-only solvable questions, augments options, and requires reading text from images, yielding substantially lower model scores of 16.8-26.9%.
-
Perception or Prejudice: Can MLLMs Go Beyond First Impressions of Personality?
Introduces the Grounded Personality Reasoning task and MM-OCEAN dataset to show that MLLMs frequently produce correct Big Five personality ratings without grounding them in observable video evidence.
-
LatentOmni: Rethinking Omni-Modal Understanding via Unified Audio-Visual Latent Reasoning
LatentOmni proposes a latent-space cross-modal reasoning framework that uses feature-level supervision and Omni-Sync Position Embedding to align and synchronize audio-visual latents, supported by a new 35K interleaved reasoning dataset and showing gains over text CoT baselines.
-
CBT-Audio: Evaluating Audio Language Models for Patient-Side Distress Intensity Estimation in CBT Session Recordings
CBT-Audio dataset shows that adding audio input improves distress intensity estimation over transcripts alone for 8 of 10 audio language models, with clearest gains when verbal content and vocal delivery diverge.
-
Omni-DuplexEval: Evaluating Real-time Duplex Omni-modal Interaction
Omni-DuplexEval creates a new benchmark and LLM-as-a-Judge framework for real-time duplex omni-modal interaction, revealing that current models score below 40% overall and struggle especially with proactive responses.
-
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.
-
EvoGround: Self-Evolving Video Agents for Video Temporal Grounding
A proposer-solver agent pair achieves supervised-level video temporal grounding and fine-grained captioning from 2.5K unlabeled videos via self-reinforcing evolution.
-
TOC-Bench: A Temporal Object Consistency Benchmark for Video Large Language Models
TOC-Bench is a new diagnostic benchmark that reveals major weaknesses in temporal object consistency for Video-LLMs, including event counting, ordering, identity reasoning, and hallucination avoidance.
-
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.
-
CGC: Compositional Grounded Contrast for Fine-Grained Multi-Image Understanding
CGC improves fine-grained multi-image understanding in MLLMs by constructing contrastive training instances from existing single-image annotations and adding a rule-based spatial reward, achieving SOTA on MIG-Bench and VLM2-Bench with transfer gains to other multimodal tasks.
-
Towards Temporal Compositional Reasoning in Long-Form Sports Videos
SportsTime benchmark and CoTR method improve multimodal AI's temporal compositional reasoning and evidence grounding in long-form sports videos.
-
Grounding Video Reasoning in Physical Signals
A new benchmark converts video clips into shared grounded event records and tests models across physics, semantic, and control prompts under original, shuffled, ablated, and masked conditions, finding selective robustness and weak spatial performance.
-
WildFireVQA: A Large-Scale Radiometric Thermal VQA Benchmark for Aerial Wildfire Monitoring
WildFireVQA is a new large-scale visual question answering benchmark that pairs RGB imagery with radiometric thermal measurements for aerial wildfire monitoring across six task categories.
-
LLM-as-Judge Framework for Evaluating Tone-Induced Hallucination in Vision-Language Models
Ghost-100 benchmark shows prompt tone drives hallucination rates and intensities in VLMs, with non-monotonic peaks at intermediate pressure and task-specific differences that aggregate metrics hide.
-
Culture-Aware Humorous Captioning: Multimodal Humor Generation across Cultural Contexts
Introduces culture-aware humorous captioning task and staged alignment framework that improves contextual fit and balances image relevance with humor in multimodal LLMs.
-
OASIS: On-Demand Hierarchical Event Memory for Streaming Video Reasoning
OASIS organizes streaming video into hierarchical events and retrieves memory on-demand via intent-driven refinement to improve long-horizon accuracy and compositional reasoning with bounded token costs.
-
MNAFT: modality neuron-aware fine-tuning of multimodal large language models for image translation
MNAFT identifies language-agnostic and language-specific neurons via activation analysis and selectively fine-tunes only relevant ones in MLLMs to close the modality gap and outperform full fine-tuning and other methods on image translation benchmarks.
-
Zero-Shot Retail Theft Detection via Orchestrated Vision Models: A Model-Agnostic, Cost-Effective Alternative to Trained Single-Model Systems
Paza is a zero-shot, model-agnostic pipeline that uses behavioral pre-filters on cheap object and pose models to trigger expensive VLMs only when needed, delivering 89.5% precision and 92.8% specificity on a synthesized shoplifting dataset at far lower cost than trained alternatives.
-
Script-a-Video: Deep Structured Audio-visual Captions via Factorized Streams and Relational Grounding
MTSS replaces monolithic video captions with factorized streams and relational grounding, yielding reported gains in understanding benchmarks and generation consistency.
-
UIPress: Bringing Optical Token Compression to UI-to-Code Generation
UIPress is the first encoder-side learned optical compression method for UI-to-Code that compresses visual tokens to 256, outperforming the uncompressed baseline by 7.5% CLIP score and the best inference-time baseline by 4.6% while delivering 9.1x TTFT speedup.
-
MARINER: A 3E-Driven Benchmark for Fine-Grained Perception and Complex Reasoning in Open-Water Environments
MARINER is a new benchmark dataset and evaluation framework for fine-grained perception and causal reasoning in open-water scenes using 16,629 images across 63 vessel categories, diverse environments, and maritime incidents.
-
VSAS-Bench: Real-Time Evaluation of Visual Streaming Assistant Models
VSAS-Bench offers temporally dense annotations and synchronous/asynchronous protocols to evaluate streaming VLMs on timeliness, consistency, accuracy, and latency trade-offs, showing that adapted conventional VLMs can outperform specialized streaming models.
-
TableVision: A Large-Scale Benchmark for Spatially Grounded Reasoning over Complex Hierarchical Tables
TableVision benchmark shows explicit spatial grounding recovers MLLM reasoning on hierarchical tables, delivering 12.3% accuracy improvement through a decoupled perception-reasoning framework.
-
OmniGUI: Benchmarking GUI Agents in Omni-Modal Smartphone Environments
OmniGUI is the first step-level benchmark supplying interleaved image, audio, and video inputs across 709 expert episodes in 29 smartphone apps to evaluate multimodal GUI agents.
-
M$^3$KG-RAG: Multi-hop Multimodal Knowledge Graph-enhanced Retrieval-Augmented Generation
M³KG-RAG improves multimodal reasoning in large language models by constructing multi-hop knowledge graphs and selectively pruning retrieved context with GRASP.
-
StreamGaze: Gaze-Guided Temporal Reasoning and Proactive Understanding in Streaming Videos
StreamGaze is a new benchmark and QA generation pipeline that measures how well MLLMs leverage gaze trajectories for temporal reasoning and proactive intention prediction in streaming egocentric videos.
-
Visual-TableQA: Open-Domain Benchmark for Reasoning over Table Images
Visual-TableQA is a new open-domain benchmark of rendered table images and complex QA pairs created via multi-LLM collaborative generation, with fine-tuned models showing robust generalization to external tests.
-
Social Human Robot Embodied Conversation (SHREC) Dataset: Benchmarking Foundational Models' Social Reasoning
SHREC is a new benchmark dataset of embodied human-robot conversations that shows substantial performance gaps in state-of-the-art foundation models on tasks involving social error detection and rationale generation.
-
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.
-
R1-VL: Learning to Reason with Multimodal Large Language Models via Step-wise Group Relative Policy Optimization
R1-VL uses StepGRPO with rule-based StepRAR and StepRVR rewards to let MLLMs learn step-by-step reasoning beyond imitation of positive paths.
-
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.
-
GME: Improving Universal Multimodal Retrieval by Multimodal LLMs
GME achieves state-of-the-art results in universal multimodal retrieval by training on a balanced synthetic multimodal dataset.
-
VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documents
VisRAG achieves 20-40% better end-to-end performance than text-based RAG by directly embedding and retrieving document images with VLMs.
-
A More Word-like Image Tokenization for MLLMs
DiVT clusters patch embeddings into coherent semantic units and adapts token count to image complexity, matching or exceeding baselines with fewer visual tokens on multimodal benchmarks.
-
Why We Look Where We Look: Emergent Human-like Fixations of a Foveated Visual Language Model Maximizing Scene Understanding
A foveated VLM trained for scene comprehension produces human-like fixations, outperforming models trained for search, classification, or with altered peripheral vision.
-
OProver: A Unified Framework for Agentic Formal Theorem Proving
OProver-32B achieves top Pass@32 scores on MiniF2F, ProverBench, and PutnamBench by combining continued pretraining with iterative agentic proving, retrieval, SFT on repairs, and RL on unresolved cases using a 6.86M-proof dataset.
-
Deep Pre-Alignment for VLMs
Deep Pre-Alignment uses a small VLM perceiver instead of ViT to pre-align visual features with LLM text space, yielding 1.9-3.0 point gains on multimodal benchmarks and 32.9% less language forgetting.
-
OmniDrop: Layer-wise Token Pruning for Omni-modal LLMs via Query-Guidance
OmniDrop is a training-free layer-wise token pruning framework for omni-modal LLMs that uses query guidance and temporal diversity to reduce prefill latency by up to 40% and memory by 14.7% while improving benchmark scores by up to 3.58 points.
-
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.
-
VideoRouter: Query-Adaptive Dual Routing for Efficient Long-Video Understanding
VideoRouter uses dual semantic and image routers for query-adaptive token compression in long-video models, delivering up to 67.9% reduction while outperforming the InternVL baseline on VideoMME, MLVU, and LongVideoBench.
-
From Priors to Perception: Grounding Video-LLMs in Physical Reality
Video-LLMs fail physical reasoning due to semantic prior dominance rather than perception deficits; a new programmatic adversarial curriculum and visual-anchored reasoning chain enable substantial gains via standard LoRA fine-tuning.
-
KARMA-MV: A Benchmark for Causal Question Answering on Music Videos
KARMA-MV is a new benchmark showing that causal knowledge graphs improve VLMs on causal audio-visual reasoning in music videos.
-
MiniCPM-o 4.5: Towards Real-Time Full-Duplex Omni-Modal Interaction
MiniCPM-o 4.5 uses the Omni-Flow streaming framework to deliver real-time full-duplex omni-modal interaction with proactive behavior in a 9B model that approaches Gemini 2.5 Flash performance.
-
See Further, Think Deeper: Advancing VLM's Reasoning Ability with Low-level Visual Cues and Reflection
ForeSight lets VLMs use low-level visual cues and mask-based visual feedback within an RL loop to reason more accurately, with the 7B model beating same-scale peers and some closed-source SOTA on a new benchmark.