Skip to content
Papers.

OmniTaskonomy: When Does Visual Generation Improve Visual Understanding? research paper by UC Berkeley, 2026

UC Berkeley · Sep 29, 2026 · Multimodal and robotics · 55 upvotes · unverified 6 days ago

Read on arXiv

What it shows

Training a model to generate visual content can encourage it to learn rich perceptual capabilities related to geometry, spatial relationships, and objectness; yet, its benefits for visual understanding remain unclear.

By Jiaxin Ge, Yiming Qin, Ji Xie and 13 more · arXiv 2609.38079 · PDF · Code

UnverifiedHugging Face's summary; not yet checked by hand.

More from UC Berkeley

All 13
PaperCitations
FreeToken: Efficient Edge-Native MoE Serving with Bandwidth-Adaptive ExecutionFreeToken is an edge-native Mixture-of-Experts serving system that dynamically maps computation and model state onto heterogeneous local hardware to run large open-weight models on personal machines.Inference and efficiency · Aug 2026 · Unverified7 weeks ago2
Playful Agentic Robot LearningEmbodied robots learn reusable skills through self-directed play and exploration, then apply these skills to improve performance on downstream tasks without additional training.Agents and evaluation · Jun 2026 · Unverified3 months ago19
Agents' Last ExamOver 1,000 real, checkable professional tasks written with 250+ industry experts; the hardest tier is far from solved.Agents and evaluation · Jun 20264 months ago21
Flash-KMeans: Fast and Memory-Efficient Exact K-MeansFlash-kmeans enables efficient online k-means clustering on GPUs through novel kernel-level optimizations that eliminate I/O bottlenecks and reduce atomic write contention.Inference and efficiency · Mar 2026 · Unverified6 months ago10
dLLM: Simple Diffusion Language ModelingA unified open-source framework is presented that standardizes core components of diffusion language modeling for reproduction, customization, and accessible development of both large and small models.Foundation models · Feb 2026 · Unverified7 months ago14
SLA2: Sparse-Linear Attention with Learnable Routing and QATSLA2 improves sparse-linear attention in diffusion models by introducing a learnable router, direct attention formulation, and quantization-aware fine-tuning for enhanced efficiency and quality.Inference and efficiency · Feb 2026 · Unverified7 months ago21
Quant VideoGen: Auto-Regressive Long Video Generation via 2-Bit KV-Cache QuantizationQuant VideoGen addresses KV cache memory limitations in autoregressive video diffusion models through semantic-aware smoothing and progressive residual quantization, achieving significant memory reduction with minimal latency impact.Inference and efficiency · Feb 2026 · Unverified8 months ago18
Residual Context Diffusion Language ModelsResidual Context Diffusion (RCD) enhances diffusion large language models by recycling discarded token information through contextual residuals, improving accuracy with minimal computational overhead.Training and scaling · Jan 2026 · Unverified8 months ago9

Other multimodal and robotics papers

Topic
PaperCitations
Segment AnythingA promptable model and a dataset of over a billion masks that cut out any object in any image.Meta · Apr 20233 years ago15.9k
BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language ModelsBLIP-2 efficiently pre-trains vision-language models using off-the-shelf frozen encoders and decoders, achieving state-of-the-art performance with fewer parameters.Salesforce · Jan 20233 years ago9,360
BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and GenerationBLIP, a Vision-Language Pre-training framework, improves performance across both understanding and generation tasks by bootstrapping captions from noisy web data, achieving state-of-the-art results on image-text retrieval, image captioning, and VQA, and showing strong generalization to video-language tasks.Salesforce · Jan 20224 years ago7,410
GPT-4o System CardGPT-4o is an omnimodal autoregressive model trained to handle text, audio, image, and video inputs, offering high-performance outputs across these modalities, with particular strengths in vision and audio.OpenAI · Oct 2024 · Unverified1 year ago5,071
SAM 3: Segment Anything with ConceptsSegment Anything Model 3 achieves state-of-the-art performance in promptable concept segmentation and tracking by leveraging a unified model architecture with decoupled recognition and localization.Meta · Nov 2025 · Unverified10 months ago1,105
DeepSeek-VL: Towards Real-World Vision-Language UnderstandingDeepSeek-VL is an open-source vision-language model that achieves state-of-the-art performance in real-world applications by combining a hybrid vision encoder with effective pretraining strategies to preserve language model capabilities.DeepSeek · Mar 2024 · Unverified2 years ago896
About this paper
Authors
Jiaxin Ge, Yiming Qin, Ji Xie and 13 more
arXiv
2609.38079 · PDF
Citations
Not counted yet · Semantic Scholar
Upvotes
55 · Hugging Face
Code
github.com/para-lost/OmniTaskonomy
Lab
UC Berkeley

Changes

What changed
New paperFound by the weekly scan, unverifiedOct 5, 2026today

New papers by email

Monday afternoons, only in weeks with new papers from the labs.

Double opt-in. Unsubscribe any time.