Papers

Filtered to visual reasoning · clear filter

Browse by term

continual learning 64reinforcement learning 34large language models 17benchmarking 12vision-language models 10generative models 8language models 8video generation 7multimodal models 6natural language processing 6robotics 6world models 6benchmarks 5diffusion models 5on-policy distillation 5policy optimization 5scalability 5self-distillation 5vision-language-action models 5autoregressive models 4computer vision 4diffusion transformers 4LLMs 4multimodal large language models 4verifiable rewards 4attention mechanisms 3embodied intelligence 3image editing 3long-term memory 3multimodal learning 3

Matching papers

See2Think: Do Multimodal Models Really Use Intermediate Visual States?

23 upvotes · 29 JUL 2026 · Siyu Yan, Zhuoran Yan, Haiying Xu et al.

This paper evaluates how multimodal large language models use intermediate visual states during reasoning and finds that these visual states are not as crucial as previously thought, but can still impact model performance under certain conditions. Practitioners might care because understanding how these models use visual states can help improve their performance and reliability.

multimodal modelsvisual reasoningintermediate statesinferencefeedbackvisual operations

Visual prompt engineering for video models

15 upvotes · 28 JUL 2026 · Robert Geirhos, Yuxuan Li, Thaddäus Wiedemer et al.

This paper explores whether visual prompt engineering, which involves modifying the task image to improve video model performance, can be as effective as text-based prompt engineering for video models, and if so, why it's worth trying for practitioners looking to improve their visual reasoning tasks.

video modelsvisual prompt engineeringimage editing modelscontinual learningvisual reasoning