University of California, Berkeley

university

https://www.berkeley.edu/

Activity Feed Request to join this org

AI & ML interests

None defined yet.

Recent Activity

rishabh2k1 submitted a paper 15 days ago

Learning, Fast and Slow: Towards LLMs That Adapt Continually

noaflynn submitted a paper about 1 month ago

COMPASS: COntinual Multilingual PEFT with Adaptive Semantic Sampling

yuezhouhu submitted a paper about 1 month ago

Speculative Decoding for Autoregressive Video Generation

View all activity

Papers

Learning, Fast and Slow: Towards LLMs That Adapt Continually

COMPASS: COntinual Multilingual PEFT with Adaptive Semantic Sampling

View all Papers

UCBerkeley 's papers 13

Submitted by

Rishabh Tiwari

Learning, Fast and Slow: Towards LLMs That Adapt Continually

UCBerkeley

University of California, Berkeley

2

Submitted by

Noah Flynn

COMPASS: COntinual Multilingual PEFT with Adaptive Semantic Sampling

UCBerkeley

University of California, Berkeley

Submitted by

Yuezhou Hu

Speculative Decoding for Autoregressive Video Generation

UCBerkeley

University of California, Berkeley

2

Submitted by

Jintao Zhang

HybridStitch: Pixel and Timestep Level Model Stitching for Diffusion Acceleration

UCBerkeley

University of California, Berkeley

2

Submitted by

Jintao Zhang

SageBwd: A Trainable Low-bit Attention

UCBerkeley

University of California, Berkeley

2

Submitted by

Haocheng Xi

Quant VideoGen: Auto-Regressive Long Video Generation via 2-Bit KV-Cache Quantization

UCBerkeley

University of California, Berkeley

Submitted by

Yuezhou Hu

Residual Context Diffusion Language Models

UCBerkeley

University of California, Berkeley

Submitted by

Jintao Zhang

TurboDiffusion: Accelerating Video Diffusion Models by 100-200 Times

UCBerkeley

University of California, Berkeley

Submitted by

Yuezhou Hu

Arbitrage: Efficient Reasoning via Advantage-Aware Speculation

UCBerkeley

University of California, Berkeley

Submitted by

taesiri

Chain-of-Visual-Thought: Teaching VLMs to See and Think Better with Continuous Visual Tokens

UCBerkeley

University of California, Berkeley

Submitted by

Junwei Yu

UnSAMv2: Self-Supervised Learning Enables Segment Anything at Any Granularity

UCBerkeley

University of California, Berkeley

Submitted by

Jiaxin Ge

Constantly Improving Image Models Need Constantly Improving Benchmarks

UCBerkeley

University of California, Berkeley

Approaching an unknown communication system by latent space exploration and causal inference

UCBerkeley

University of California, Berkeley