mlsanityData to 7 Oct 2026

Multimodal LLMs / vision-language models

16 directions · 12,360 papers · 58 this week

Busiest this week

Most new since Mon 5 Oct

MLLM jailbreaks and adversarial robustness

850 papers9 this weekHigh confidence in this grouping: the papers form one clear topicHigh confidence in this grouping: the papers form one clear topic · About

  1. Detecting Adversarial Images through Response Profiles of Vision-Language Models
  2. Open-MMUnlearning: Unifying Methods and Evaluation for MLLM Unlearning
  3. Purifying Backdoored Large Vision-Language Models by Removing Hijacked Directions
  4. Understanding and Mitigating Token-Pruning-Induced Vulnerabilities in VLMs
  5. It Is Not Seeing the Hazard: A Frozen Vision-Language Safety Score Measures Its Caption Bank
Open →

All directions

Largest first · week from Mon 5 Oct

  1. Embodied and 3D vision-language modelsreasoning, spatial, vlms, languageLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About2,2888Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  2. VLM benchmarks and visual reasoningvisual, reasoning, vlms, vision-languageLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About2,0635Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  3. MLLM training and visual instruction tuningmultimodal, mllms, language, largeLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About1,4683Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  4. Prompt learning and adaptation of CLIPprompt, clip, vision-language, adaptationHigh confidence in this grouping: the papers form one clear topic · About1,0457High confidence in this grouping: the papers form one clear topic
  5. Medical vision-language modelsmedical, clinical, vision-language, imageHigh confidence in this grouping: the papers form one clear topic · About9476High confidence in this grouping: the papers form one clear topic
  6. MLLM jailbreaks and adversarial robustnessadversarial, safety, attacks, attackHigh confidence in this grouping: the papers form one clear topic · About8509High confidence in this grouping: the papers form one clear topic
  7. Video understanding with MLLMsvideo, understanding, temporal, video understandingHigh confidence in this grouping: the papers form one clear topic · About7422High confidence in this grouping: the papers form one clear topic
  8. Visual token pruning and efficiencytokens, token, pruning, visualHigh confidence in this grouping: the papers form one clear topic · About6355High confidence in this grouping: the papers form one clear topic
  9. Hallucination mitigation in LVLMshallucination, hallucinations, visual, lvlmsHigh confidence in this grouping: the papers form one clear topic · About5992High confidence in this grouping: the papers form one clear topic
  10. Document understanding and multimodal RAGretrieval, document, multimodal, visualMedium confidence in this grouping: mostly one topic, with blurry edges · About4800Medium confidence in this grouping: mostly one topic, with blurry edges
  11. Chart understanding and anomaly detectionchart, anomaly, reasoning, visualLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About3963Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  12. Driving VLMs and open-vocabulary detectiondriving, autonomous, autonomous driving, detectionMedium confidence in this grouping: mostly one topic, with blurry edges · About3901Medium confidence in this grouping: mostly one topic, with blurry edges
  13. Image editing and quality assessmentediting, quality, image, diffusionLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · About2021Low confidence in this grouping: the papers may belong to more than one topic, or the group may change
  14. Social bias and fairness in VLMsbias, gender, biases, socialMedium confidence in this grouping: mostly one topic, with blurry edges · About1230Medium confidence in this grouping: mostly one topic, with blurry edges
  15. Multimodal emotion understandingemotion, emotional, emotions, affectiveHigh confidence in this grouping: the papers form one clear topic · About900High confidence in this grouping: the papers form one clear topic
  16. Agricultural vision-language modelsagricultural, plant, agriculture, speciesHigh confidence in this grouping: the papers form one clear topic · About420High confidence in this grouping: the papers form one clear topic
6 new this week aren’t in a direction yet
  1. End-to-End Autonomous Generation of Human Assembly Plans
  2. Adversarial Images Hijack Web Agents from Visual Grounding to Browser Execution
  3. RSJEV: Discriminative Remote Sensing Scene Classification with Multimodal Large Language Models
  4. A doctrine-grounded visual question answering dataset for Tactical Combat Casualty Care
  5. Demo: Vision-Language Model-Guided Online Calibration of an Electromagnetic Digital Twin
  6. Benchmarking Jailbreak Guardrails for Embodied Agents

Weekly activity

Papers per week, 22 Jun – 5 Oct (last week partial)

  1. DirectionConfidencepapers16 wks
  2. Embodied and 3D vision-language modelsLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 25; 29 Jun: 27; 6 Jul: 15; 13 Jul: 31; 20 Jul: 21; 27 Jul: 18; 3 Aug: 27; 10 Aug: 20; 17 Aug: 22; 24 Aug: 20; 31 Aug: 21; 7 Sep: 24; 14 Sep: 21; 21 Sep: 25; 28 Sep: 45; 5 Oct: 82,288370
  3. VLM benchmarks and visual reasoningLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 20; 29 Jun: 26; 6 Jul: 16; 13 Jul: 15; 20 Jul: 10; 27 Jul: 20; 3 Aug: 25; 10 Aug: 18; 17 Aug: 12; 24 Aug: 24; 31 Aug: 26; 7 Sep: 9; 14 Sep: 18; 21 Sep: 22; 28 Sep: 32; 5 Oct: 52,063298
  4. Visual token pruning and efficiencyHigh confidence in this grouping: the papers form one clear topic · AboutHigh confidence in this grouping: the papers form one clear topicPer week: 22 Jun: 5; 29 Jun: 5; 6 Jul: 7; 13 Jul: 4; 20 Jul: 6; 27 Jul: 9; 3 Aug: 19; 10 Aug: 6; 17 Aug: 3; 24 Aug: 12; 31 Aug: 11; 7 Sep: 3; 14 Sep: 9; 21 Sep: 8; 28 Sep: 25; 5 Oct: 5635137
  5. Medical vision-language modelsHigh confidence in this grouping: the papers form one clear topic · AboutHigh confidence in this grouping: the papers form one clear topicPer week: 22 Jun: 14; 29 Jun: 9; 6 Jul: 4; 13 Jul: 13; 20 Jul: 10; 27 Jul: 18; 3 Aug: 8; 10 Aug: 12; 17 Aug: 6; 24 Aug: 7; 31 Aug: 9; 7 Sep: 3; 14 Sep: 8; 21 Sep: 12; 28 Sep: 12; 5 Oct: 6947151
  6. Prompt learning and adaptation of CLIPHigh confidence in this grouping: the papers form one clear topic · AboutHigh confidence in this grouping: the papers form one clear topicPer week: 22 Jun: 8; 29 Jun: 6; 6 Jul: 5; 13 Jul: 4; 20 Jul: 5; 27 Jul: 3; 3 Aug: 5; 10 Aug: 3; 17 Aug: 7; 24 Aug: 10; 31 Aug: 3; 7 Sep: 5; 14 Sep: 5; 21 Sep: 9; 28 Sep: 10; 5 Oct: 71,04595
  7. Video understanding with MLLMsHigh confidence in this grouping: the papers form one clear topic · AboutHigh confidence in this grouping: the papers form one clear topicPer week: 22 Jun: 10; 29 Jun: 12; 6 Jul: 1; 13 Jul: 4; 20 Jul: 2; 27 Jul: 5; 3 Aug: 6; 10 Aug: 11; 17 Aug: 5; 24 Aug: 11; 31 Aug: 7; 7 Sep: 2; 14 Sep: 4; 21 Sep: 6; 28 Sep: 19; 5 Oct: 2742107
  8. MLLM jailbreaks and adversarial robustnessHigh confidence in this grouping: the papers form one clear topic · AboutHigh confidence in this grouping: the papers form one clear topicPer week: 22 Jun: 8; 29 Jun: 7; 6 Jul: 8; 13 Jul: 5; 20 Jul: 5; 27 Jul: 10; 3 Aug: 10; 10 Aug: 7; 17 Aug: 13; 24 Aug: 5; 31 Aug: 12; 7 Sep: 4; 14 Sep: 3; 21 Sep: 3; 28 Sep: 12; 5 Oct: 9850121
  9. MLLM training and visual instruction tuningLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 16; 29 Jun: 13; 6 Jul: 3; 13 Jul: 3; 20 Jul: 3; 27 Jul: 17; 3 Aug: 9; 10 Aug: 9; 17 Aug: 7; 24 Aug: 12; 31 Aug: 14; 7 Sep: 6; 14 Sep: 3; 21 Sep: 6; 28 Sep: 11; 5 Oct: 31,468135
  10. Document understanding and multimodal RAGMedium confidence in this grouping: mostly one topic, with blurry edges · AboutMedium confidence in this grouping: mostly one topic, with blurry edgesPer week: 22 Jun: 6; 29 Jun: 1; 6 Jul: 6; 13 Jul: 1; 20 Jul: 2; 27 Jul: 3; 3 Aug: 8; 10 Aug: 1; 17 Aug: 2; 24 Aug: 10; 31 Aug: 6; 7 Sep: 1; 14 Sep: 3; 21 Sep: 4; 28 Sep: 11; 5 Oct: 048065
  11. Chart understanding and anomaly detectionLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 1; 29 Jun: 7; 6 Jul: 3; 13 Jul: 2; 20 Jul: 3; 27 Jul: 4; 3 Aug: 7; 10 Aug: 7; 17 Aug: 2; 24 Aug: 5; 31 Aug: 6; 7 Sep: 8; 14 Sep: 3; 21 Sep: 2; 28 Sep: 5; 5 Oct: 339668
  12. Hallucination mitigation in LVLMsHigh confidence in this grouping: the papers form one clear topic · AboutHigh confidence in this grouping: the papers form one clear topicPer week: 22 Jun: 5; 29 Jun: 8; 6 Jul: 1; 13 Jul: 3; 20 Jul: 3; 27 Jul: 7; 3 Aug: 7; 10 Aug: 7; 17 Aug: 4; 24 Aug: 7; 31 Aug: 8; 7 Sep: 3; 14 Sep: 3; 21 Sep: 3; 28 Sep: 4; 5 Oct: 259975
  13. Driving VLMs and open-vocabulary detectionMedium confidence in this grouping: mostly one topic, with blurry edges · AboutMedium confidence in this grouping: mostly one topic, with blurry edgesPer week: 22 Jun: 4; 29 Jun: 5; 6 Jul: 3; 13 Jul: 3; 20 Jul: 2; 27 Jul: 3; 3 Aug: 4; 10 Aug: 3; 17 Aug: 0; 24 Aug: 2; 31 Aug: 2; 7 Sep: 2; 14 Sep: 2; 21 Sep: 3; 28 Sep: 6; 5 Oct: 139045
  14. Image editing and quality assessmentLow confidence in this grouping: the papers may belong to more than one topic, or the group may change · AboutLow confidence in this grouping: the papers may belong to more than one topic, or the group may changePer week: 22 Jun: 0; 29 Jun: 3; 6 Jul: 0; 13 Jul: 1; 20 Jul: 1; 27 Jul: 1; 3 Aug: 1; 10 Aug: 1; 17 Aug: 3; 24 Aug: 2; 31 Aug: 1; 7 Sep: 1; 14 Sep: 2; 21 Sep: 0; 28 Sep: 4; 5 Oct: 120222
  15. Social bias and fairness in VLMsMedium confidence in this grouping: mostly one topic, with blurry edges · AboutMedium confidence in this grouping: mostly one topic, with blurry edgesPer week: 22 Jun: 1; 29 Jun: 1; 6 Jul: 0; 13 Jul: 1; 20 Jul: 0; 27 Jul: 0; 3 Aug: 0; 10 Aug: 2; 17 Aug: 0; 24 Aug: 2; 31 Aug: 3; 7 Sep: 1; 14 Sep: 2; 21 Sep: 2; 28 Sep: 2; 5 Oct: 012317
  16. Multimodal emotion understandingHigh confidence in this grouping: the papers form one clear topic · AboutHigh confidence in this grouping: the papers form one clear topicPer week: 22 Jun: 0; 29 Jun: 0; 6 Jul: 0; 13 Jul: 1; 20 Jul: 1; 27 Jul: 0; 3 Aug: 2; 10 Aug: 2; 17 Aug: 0; 24 Aug: 3; 31 Aug: 0; 7 Sep: 0; 14 Sep: 0; 21 Sep: 0; 28 Sep: 3; 5 Oct: 09012
  17. Agricultural vision-language modelsHigh confidence in this grouping: the papers form one clear topic · AboutHigh confidence in this grouping: the papers form one clear topicPer week: 22 Jun: 0; 29 Jun: 1; 6 Jul: 0; 13 Jul: 0; 20 Jul: 0; 27 Jul: 0; 3 Aug: 1; 10 Aug: 0; 17 Aug: 1; 24 Aug: 1; 31 Aug: 1; 7 Sep: 2; 14 Sep: 0; 21 Sep: 0; 28 Sep: 1; 5 Oct: 0428