Picture for Ion Stoica

Ion Stoica

Why Do Multi-Agent LLM Systems Fail?

Add code
Mar 17, 2025
Viaarxiv icon

WorldModelBench: Judging Video Generation Models As World Models

Add code
Feb 28, 2025
Viaarxiv icon

Prompt-to-Leaderboard

Add code
Feb 20, 2025
Viaarxiv icon

Optimizing Model Selection for Compound AI Systems

Add code
Feb 20, 2025
Viaarxiv icon

S*: Test Time Scaling for Code Generation

Add code
Feb 20, 2025
Viaarxiv icon

Autellix: An Efficient Serving Engine for LLM Agents as General Programs

Add code
Feb 19, 2025
Viaarxiv icon

The Danger of Overthinking: Examining the Reasoning-Action Dilemma in Agentic Tasks

Add code
Feb 12, 2025
Viaarxiv icon

LLMs Can Easily Learn to Reason from Demonstrations Structure, not content, is what matters!

Add code
Feb 11, 2025
Viaarxiv icon

Efficient-vDiT: Efficient Video Diffusion Transformers With Attention Tile

Add code
Feb 10, 2025
Viaarxiv icon

Twilight: Adaptive Attention Sparsity with Hierarchical Top-$p$ Pruning

Add code
Feb 06, 2025
Figure 1 for Twilight: Adaptive Attention Sparsity with Hierarchical Top-$p$ Pruning
Figure 2 for Twilight: Adaptive Attention Sparsity with Hierarchical Top-$p$ Pruning
Figure 3 for Twilight: Adaptive Attention Sparsity with Hierarchical Top-$p$ Pruning
Figure 4 for Twilight: Adaptive Attention Sparsity with Hierarchical Top-$p$ Pruning
Viaarxiv icon