Voozh

AI & ML interests

None defined yet.

Recent Activity

👁 Image

changdae submitted a paper 3 days ago

Neglected Free Lunch from Post-training: Progress Advantage for LLM Agents

👁 Image

zhuomingliu submitted a paper 18 days ago

DRIFT: A Residual Flow Adapter for Decoding Continuous Outputs in Vision-Language Models

👁 Image

Mercury7353 authored a paper 27 days ago

When Does Multi-Agent RL Improve LLM Workflows? Workflow, Scale, and Policy-Sharing Tradeoffs

View all activity

Papers

👁 Image

Neglected Free Lunch from Post-training: Progress Advantage for LLM Agents

👁 Image

DRIFT: A Residual Flow Adapter for Decoding Continuous Outputs in Vision-Language Models

View all Papers

👁 Image

Submitted by

👁 Image

Changdae Oh

Neglected Free Lunch from Post-training: Progress Advantage for LLM Agents

👁 uw-madison
University of Wisconsin - Madison

7 1

Submitted by

👁 Image

Zhuoming Liu

DRIFT: A Residual Flow Adapter for Decoding Continuous Outputs in Vision-Language Models

👁 uw-madison
University of Wisconsin - Madison

3 2

👁 Image

Submitted by

👁 Image

Anirudh Sundara Rajan

From Plans to Pixels: Learning to Plan and Orchestrate for Open-Ended Image Editing

👁 uw-madison
University of Wisconsin - Madison

👁 Image

Submitted by

👁 Image

Jaden Park

Exploration and Exploitation Errors Are Measurable for Language Model Agents

👁 uw-madison
University of Wisconsin - Madison

7 3

👁 Image

Submitted by

👁 Image

Gabe Orlanski

SlopCodeBench: Benchmarking How Coding Agents Degrade Over Long-Horizon Iterative Tasks

👁 uw-madison
University of Wisconsin - Madison

81 2

👁 Image

Submitted by

👁 Image

Tengyang Xie

Understanding Behavior Cloning with Action Quantization

👁 uw-madison
University of Wisconsin - Madison

👁 Image

Submitted by

👁 Image

Tengyang Xie

Breaking the Capability Ceiling of LLM Post-Training by Reintroducing Markov States

👁 uw-madison
University of Wisconsin - Madison

👁 Image

Submitted by

👁 Image

Jiayu (Mila) Wang

SkillOrchestra: Learning to Route Agents via Skill Transfer

👁 uw-madison
University of Wisconsin - Madison

66 6

👁 Image

Submitted by

👁 Image

Jongwon Jeong

TAPE: Tool-Guided Adaptive Planning and Constrained Execution in Language Model Agents

👁 uw-madison
University of Wisconsin - Madison

7 2

👁 Image

Submitted by

👁 Image

Changdae Oh

Thinking Makes LLM Agents Introverted: How Mandatory Thinking Can Backfire in User-Engaged Agents

👁 uw-madison
University of Wisconsin - Madison

👁 Image

Submitted by

👁 Image

Changdae Oh

Towards Reducible Uncertainty Modeling for Reliable Large Language Model Agents

👁 uw-madison
University of Wisconsin - Madison

👁 Image

Submitted by

👁 Image

Mu Cai

Contamination Detection for VLMs using Multi-Modal Semantic Perturbation

👁 uw-madison
University of Wisconsin - Madison

👁 Image

UniTalk: Towards Universal Active Speaker Detection in Real World Scenarios

👁 uw-madison
University of Wisconsin - Madison

👁 Image

VersaPRM: Multi-Domain Process Reward Model via Synthetic Reasoning Data

👁 uw-madison
University of Wisconsin - Madison

URL: https://huggingface.co/uw-madison/papers

⇱ uw-madison (University of Wisconsin - Madison)

AI & ML interests

Recent Activity

Papers

Neglected Free Lunch from Post-training: Progress Advantage for LLM Agents

DRIFT: A Residual Flow Adapter for Decoding Continuous Outputs in Vision-Language Models

From Plans to Pixels: Learning to Plan and Orchestrate for Open-Ended Image Editing

Exploration and Exploitation Errors Are Measurable for Language Model Agents

SlopCodeBench: Benchmarking How Coding Agents Degrade Over Long-Horizon Iterative Tasks

Understanding Behavior Cloning with Action Quantization

Breaking the Capability Ceiling of LLM Post-Training by Reintroducing Markov States

SkillOrchestra: Learning to Route Agents via Skill Transfer

TAPE: Tool-Guided Adaptive Planning and Constrained Execution in Language Model Agents

Thinking Makes LLM Agents Introverted: How Mandatory Thinking Can Backfire in User-Engaged Agents

Towards Reducible Uncertainty Modeling for Reliable Large Language Model Agents

Contamination Detection for VLMs using Multi-Modal Semantic Perturbation

UniTalk: Towards Universal Active Speaker Detection in Real World Scenarios

VersaPRM: Multi-Domain Process Reward Model via Synthetic Reasoning Data