dulacp (Pierre Dulac)

upvoted an article 3 months ago

Article

Introducing ColQwen-Omni: Retrieve in every modality

Jul 17

•

75

upvoted a paper 6 months ago

Evaluation is All You Need: Strategic Overclaiming of LLM Reasoning Capabilities Through Evaluation Design

Paper • 2506.04734 • Published Jun 5 • 20

upvoted a paper 7 months ago

BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset

Paper • 2505.09568 • Published May 14 • 97

upvoted a paper 8 months ago

Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?

Paper • 2504.13837 • Published Apr 18 • 138

upvoted 3 papers 10 months ago

Superintelligent Agents Pose Catastrophic Risks: Can Scientist AI Offer a Safer Path?

Paper • 2502.15657 • Published Feb 21 • 5

Magma: A Foundation Model for Multimodal AI Agents

Paper • 2502.13130 • Published Feb 18 • 58

Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling

Paper • 2412.05271 • Published Dec 6, 2024 • 159

upvoted an article 10 months ago

Article

Open R1: Update #2

Feb 10

•

218

upvoted a collection 10 months ago

🤖 Agents

Collection

21 items • Updated Dec 31, 2024 • 169

upvoted 2 articles 10 months ago

Article

Introducing smolagents: simple agents that write actions in code.

+1

Dec 31, 2024

•

1.15k

Article

Open-R1: a fully open reproduction of DeepSeek-R1

+1

Jan 28

•

887

upvoted 3 papers 11 months ago

DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning

Paper • 2501.12948 • Published Jan 22 • 429

1.58-bit FLUX

Paper • 2412.18653 • Published Dec 24, 2024 • 85

Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs

Paper • 2412.21187 • Published Dec 30, 2024 • 40

upvoted a collection 11 months ago

PixMo

Collection

A set of vision-language datasets built by Ai2 and used to train the Molmo family of models. Read more at https://molmo.allenai.org/blog • 10 items • Updated 7 days ago • 81

upvoted a paper 11 months ago

Mulberry: Empowering MLLM with o1-like Reasoning and Reflection via Collective Monte Carlo Tree Search

Paper • 2412.18319 • Published Dec 24, 2024 • 39

upvoted 4 papers 12 months ago

Pierre Dulac

AI & ML interests

Organizations

Introducing ColQwen-Omni: Retrieve in every modality

Evaluation is All You Need: Strategic Overclaiming of LLM Reasoning Capabilities Through Evaluation Design

BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset

Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?

Superintelligent Agents Pose Catastrophic Risks: Can Scientist AI Offer a Safer Path?

Magma: A Foundation Model for Multimodal AI Agents

Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling

Open R1: Update #2

🤖 Agents

Introducing smolagents: simple agents that write actions in code.

Open-R1: a fully open reproduction of DeepSeek-R1

DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning

1.58-bit FLUX

Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs

PixMo

Mulberry: Empowering MLLM with o1-like Reasoning and Reflection via Collective Monte Carlo Tree Search

Offline Reinforcement Learning for LLM Multi-Step Reasoning

Phi-4 Technical Report

PaliGemma 2: A Family of Versatile VLMs for Transfer

SNOOPI: Supercharged One-step Diffusion Distillation with Proper Guidance

Pierre Dulac

AI & ML interests

Organizations

dulacp's activity

Introducing ColQwen-Omni: Retrieve in every modality

Open R1: Update #2

Introducing smolagents: simple agents that write actions in code.

Open-R1: a fully open reproduction of DeepSeek-R1