Interesting work on agentic image generation. Co-train the generator and agent to optimize the system. https://haozheh3.github.io/SearchGen/ #imagegeneration #agenticai #multimodalllm (via https://message.haoxiang.org)
August 25, 2026 at 7:31 PM
Interesting work on agentic image generation. Co-train the generator and agent to optimize the system. https://haozheh3.github.io/SearchGen/ #imagegeneration #agenticai #multimodalllm (via https://message.haoxiang.org)
Researchers fine‑tuned a multimodal LLM so household robots can ask clarification questions, boosting task success by 10.4‑16.5% over baselines. Posted 1 Apr 2025. Read more: https://getnews.me/multimodal-llms-learn-to-ask-clarifying-questions-for-household-robots/ #multimodalllm #householdrobots
October 6, 2025 at 1:31 PM
Researchers fine‑tuned a multimodal LLM so household robots can ask clarification questions, boosting task success by 10.4‑16.5% over baselines. Posted 1 Apr 2025. Read more: https://getnews.me/multimodal-llms-learn-to-ask-clarifying-questions-for-household-robots/ #multimodalllm #householdrobots
RL‑fine‑tuned multimodal LLMs improve Ask-to-Act benchmark success by 10‑16% over baselines, learning to ask minimal clarification questions without human‑provided rewards. Read more: https://getnews.me/fine-tuning-multimodal-llms-for-embodied-agents-that-ask-questions/ #multimodalllm #embodiedai
October 3, 2025 at 7:31 AM
RL‑fine‑tuned multimodal LLMs improve Ask-to-Act benchmark success by 10‑16% over baselines, learning to ask minimal clarification questions without human‑provided rewards. Read more: https://getnews.me/fine-tuning-multimodal-llms-for-embodied-agents-that-ask-questions/ #multimodalllm #embodiedai
GHOST generates images that cause multimodal LLMs to hallucinate missing objects, achieving a success rate over 28% versus ~1% for prior methods. The cues also fooled GPT‑4o at 66.5%. Read more: https://getnews.me/ghost-images-that-trigger-hallucinations-in-multimodal-llms/ #multimodalllm #ghost
October 1, 2025 at 3:56 AM
GHOST generates images that cause multimodal LLMs to hallucinate missing objects, achieving a success rate over 28% versus ~1% for prior methods. The cues also fooled GPT‑4o at 66.5%. Read more: https://getnews.me/ghost-images-that-trigger-hallucinations-in-multimodal-llms/ #multimodalllm #ghost
Vision Function Layers (VFLs) in multimodal LLMs concentrate visual tasks into a few decoder layers; VFL‑select keeps 98 % performance using only ~20 % of the data. Read more: https://getnews.me/identifying-vision-function-layers-in-multimodal-large-language-models/ #vfl #multimodalllm
September 30, 2025 at 11:57 PM
Vision Function Layers (VFLs) in multimodal LLMs concentrate visual tasks into a few decoder layers; VFL‑select keeps 98 % performance using only ~20 % of the data. Read more: https://getnews.me/identifying-vision-function-layers-in-multimodal-large-language-models/ #vfl #multimodalllm
A 35‑page survey shows multimodal LLMs beat text‑only baselines on emotion classification; instruction‑tuned models achieve high generalization. Read more: https://getnews.me/survey-highlights-advances-in-multimodal-large-language-models-for-emotion-recognition/ #multimodalllm #emotionrecognition
September 30, 2025 at 8:03 PM
A 35‑page survey shows multimodal LLMs beat text‑only baselines on emotion classification; instruction‑tuned models achieve high generalization. Read more: https://getnews.me/survey-highlights-advances-in-multimodal-large-language-models-for-emotion-recognition/ #multimodalllm #emotionrecognition
A new study fine‑tuned a multimodal LLM with a dataset on weight, stability, size and footprint, letting a humanoid robot plan stacking tasks and beat a baseline model in simulations. https://getnews.me/multimodal-llms-enable-preference-based-long-horizon-robotic-stacking/ #multimodalllm #robotics
September 30, 2025 at 6:20 PM
A new study fine‑tuned a multimodal LLM with a dataset on weight, stability, size and footprint, letting a humanoid robot plan stacking tasks and beat a baseline model in simulations. https://getnews.me/multimodal-llms-enable-preference-based-long-horizon-robotic-stacking/ #multimodalllm #robotics
Removing certain vision encoders can boost accuracy by up to 3.6%, while using just one or two encoders retains over 90% of baseline performance on most non‑OCR tasks. Read more: https://getnews.me/multimodal-llms-reveal-redundancy-in-multiple-vision-encoders/ #multimodalllm #visionencoder #ai
September 29, 2025 at 11:18 PM
Removing certain vision encoders can boost accuracy by up to 3.6%, while using just one or two encoders retains over 90% of baseline performance on most non‑OCR tasks. Read more: https://getnews.me/multimodal-llms-reveal-redundancy-in-multiple-vision-encoders/ #multimodalllm #visionencoder #ai
Study shows household robots can ask clarification questions using LLMs fine‑tuned with reinforcement learning, significantly boosting performance by 10.4%–16.5%. Read more: https://getnews.me/multimodal-llms-empower-household-robots-to-ask-clarifying-questions/ #multimodalllm #householdrobots
September 29, 2025 at 8:35 PM
Study shows household robots can ask clarification questions using LLMs fine‑tuned with reinforcement learning, significantly boosting performance by 10.4%–16.5%. Read more: https://getnews.me/multimodal-llms-empower-household-robots-to-ask-clarifying-questions/ #multimodalllm #householdrobots
WebRSSBench, a benchmark for multimodal LLMs, defines eight web tasks and evaluated twelve models, exposing gaps in compositional reasoning and reduced robustness to layout changes. https://getnews.me/benchmark-tests-mllm-web-understanding-reasoning-robustness-safety/ #webrssbench #multimodalllm
September 29, 2025 at 8:41 AM
WebRSSBench, a benchmark for multimodal LLMs, defines eight web tasks and evaluated twelve models, exposing gaps in compositional reasoning and reduced robustness to layout changes. https://getnews.me/benchmark-tests-mllm-web-understanding-reasoning-robustness-safety/ #webrssbench #multimodalllm
MM‑Detect reveals text and image leakage in multimodal LLMs, detecting contamination in 12 models across five benchmarks; paper updates run through 20 Sep 2025. Read more: https://getnews.me/systematic-study-finds-text-and-image-leakage-in-multimodal-llms/ #mmdetect #multimodalllm
September 25, 2025 at 2:08 PM
MM‑Detect reveals text and image leakage in multimodal LLMs, detecting contamination in 12 models across five benchmarks; paper updates run through 20 Sep 2025. Read more: https://getnews.me/systematic-study-finds-text-and-image-leakage-in-multimodal-llms/ #mmdetect #multimodalllm
A new multimodal LLM framework uses Chain‑of‑Thought prompting to evaluate edited audio, giving text explanations that align with human MOS ratings. Code is on GitHub. https://getnews.me/interpretable-audio-editing-evaluation-with-chain-of-thought-llms/ #audioediting #multimodalllm
September 24, 2025 at 4:07 PM
A new multimodal LLM framework uses Chain‑of‑Thought prompting to evaluate edited audio, giving text explanations that align with human MOS ratings. Code is on GitHub. https://getnews.me/interpretable-audio-editing-evaluation-with-chain-of-thought-llms/ #audioediting #multimodalllm
The MCP framework boosts multimodal LLM accuracy by 15‑30% and cuts compute time by ~40%, while its Presenter layer reaches 90% of human‑rated interpretability. Read more: https://getnews.me/control-theoretic-framework-improves-multimodal-llm-efficiency/ #multimodalllm #controltheory #efficiency
September 24, 2025 at 12:00 PM
The MCP framework boosts multimodal LLM accuracy by 15‑30% and cuts compute time by ~40%, while its Presenter layer reaches 90% of human‑rated interpretability. Read more: https://getnews.me/control-theoretic-framework-improves-multimodal-llm-efficiency/ #multimodalllm #controltheory #efficiency
LIR‑GAD adds and tokens to a multimodal LLM, boosting accuracy and interpretability for group activity detection on standard benchmarks. Read more: https://getnews.me/language-instructed-reasoning-improves-group-activity-detection/ #groupactivitydetection #multimodalllm
September 22, 2025 at 3:38 PM
LIR‑GAD adds and tokens to a multimodal LLM, boosting accuracy and interpretability for group activity detection on standard benchmarks. Read more: https://getnews.me/language-instructed-reasoning-improves-group-activity-detection/ #groupactivitydetection #multimodalllm
The 16 September 2025 study finds multimodal LLMs detect artifacts and style but often miss anatomical accuracy, unlike humans who reliably judge all six quality attributes. Read more: https://getnews.me/examining-how-humans-and-multimodal-llms-judge-generated-images/ #multimodalllm #imageevaluation
September 18, 2025 at 10:53 AM
The 16 September 2025 study finds multimodal LLMs detect artifacts and style but often miss anatomical accuracy, unlike humans who reliably judge all six quality attributes. Read more: https://getnews.me/examining-how-humans-and-multimodal-llms-judge-generated-images/ #multimodalllm #imageevaluation
ANTS, a training‑free, zero‑shot method using multimodal LLMs, cut the false‑positive rate by 4.2% at 95% recall on ImageNet. The paper was released September 11 2025. Read more: https://getnews.me/ants-method-uses-multimodal-llms-to-boost-ood-detection/ #ants #ooddetection #multimodalllm
September 17, 2025 at 1:41 PM
ANTS, a training‑free, zero‑shot method using multimodal LLMs, cut the false‑positive rate by 4.2% at 95% recall on ImageNet. The paper was released September 11 2025. Read more: https://getnews.me/ants-method-uses-multimodal-llms-to-boost-ood-detection/ #ants #ooddetection #multimodalllm
Imagine an AI platform that talks, sees, and writes—all in one, built for enterprise scale and fully tunable via APIs. The future of multimodal assistants is here. Dive into the details! #EnterpriseAI #MultimodalLLM #APICustomization
🔗 aidailypost.com/news/enterpr...
🔗 aidailypost.com/news/enterpr...
May 16, 2026 at 6:08 PM
Imagine an AI platform that talks, sees, and writes—all in one, built for enterprise scale and fully tunable via APIs. The future of multimodal assistants is here. Dive into the details! #EnterpriseAI #MultimodalLLM #APICustomization
🔗 aidailypost.com/news/enterpr...
🔗 aidailypost.com/news/enterpr...
Z.AI just dropped GLM‑4.7, an open‑source LLM that cranks up coding, reasoning, and text‑vision with massive context windows and a sleek API. Looks like a serious Claude challenger. Dive in for the details! #GLM47 #OpenSourceAI #MultimodalLLM
🔗 aidailypost.com/news/zai-rel...
🔗 aidailypost.com/news/zai-rel...
December 25, 2025 at 9:11 AM
Z.AI just dropped GLM‑4.7, an open‑source LLM that cranks up coding, reasoning, and text‑vision with massive context windows and a sleek API. Looks like a serious Claude challenger. Dive in for the details! #GLM47 #OpenSourceAI #MultimodalLLM
🔗 aidailypost.com/news/zai-rel...
🔗 aidailypost.com/news/zai-rel...
This #CVIA study presents the first #MultimodalLLM driven preoperative planning system for #TAVR, integrating #Imaging and #ClinicalData to generate structured surgical strategies and risk analysis reports for precision #CardiovascularCare.
#OpenAccess cvia-journal.org/tavr-multimo...
#OpenAccess cvia-journal.org/tavr-multimo...
April 10, 2026 at 1:30 PM
This #CVIA study presents the first #MultimodalLLM driven preoperative planning system for #TAVR, integrating #Imaging and #ClinicalData to generate structured surgical strategies and risk analysis reports for precision #CardiovascularCare.
#OpenAccess cvia-journal.org/tavr-multimo...
#OpenAccess cvia-journal.org/tavr-multimo...
I'll be giving an invited talk at the DaQuaMRec workshop at #RecSys2026 on September 28 in Minneapolis, MN! 🇺🇸
Many thanks to the organizers. Looking forward to it!
More information:
sites.google.com/view/daquamr...
#RecSys #MultimodalLLM #LLM @recsys.bsky.social
Many thanks to the organizers. Looking forward to it!
More information:
sites.google.com/view/daquamr...
#RecSys #MultimodalLLM #LLM @recsys.bsky.social
July 31, 2026 at 8:46 AM
I'll be giving an invited talk at the DaQuaMRec workshop at #RecSys2026 on September 28 in Minneapolis, MN! 🇺🇸
Many thanks to the organizers. Looking forward to it!
More information:
sites.google.com/view/daquamr...
#RecSys #MultimodalLLM #LLM @recsys.bsky.social
Many thanks to the organizers. Looking forward to it!
More information:
sites.google.com/view/daquamr...
#RecSys #MultimodalLLM #LLM @recsys.bsky.social