SynthLabs

company

Verified

https://www.SynthLabs.ai

Activity Feed Request to join this org

AI & ML interests

Scaling up good synthetic reasoning. Post-training and synthetic data research lab.

Recent Activity

violetxi submitted a paper 10 days ago

ExpRL: Exploratory RL for LLM Mid-Training

Asap7772 authored a paper 9 months ago

Personalized Preference Fine-tuning of Diffusion Models

Asap7772 authored a paper 9 months ago

Big-Math: A Large-Scale, High-Quality Math Dataset for Reinforcement Learning in Language Models

View all activity

submitted a paper to Daily Papers 10 days ago

ExpRL: Exploratory RL for LLM Mid-Training

Paper • 2606.17024 • Published 11 days ago • 5

authored 3 papers 9 months ago

Personalized Preference Fine-tuning of Diffusion Models

Paper • 2501.06655 • Published Jan 11, 2025 • 1

Big-Math: A Large-Scale, High-Quality Math Dataset for Reinforcement Learning in Language Models

Paper • 2502.17387 • Published Feb 24, 2025 • 8

RLAD: Training LLMs to Discover Abstractions for Solving Reasoning Problems

Paper • 2510.02263 • Published Oct 2, 2025 • 9

updated a Space 10 months ago

README

published 2 models about 1 year ago

SynthLabsAI/ALP_R1_Qwen1.5B

Reinforcement Learning • 2B • Updated Jun 24, 2025 • 4

SynthLabsAI/ALP_DeepScaleR_1.5B_C16K

Reinforcement Learning • 2B • Updated Jun 24, 2025 • 4 • 3

updated a collection about 1 year ago

Adaptive Length Penalty

Teaching language models to think efficiently with Adaptive Length Penalty (ALP) • 3 items • Updated Jun 24, 2025 • 1

updated 2 models about 1 year ago

SynthLabsAI/ALP_DeepScaleR_1.5B_C16K

Reinforcement Learning • 2B • Updated Jun 24, 2025 • 4 • 3

SynthLabsAI/ALP_R1_Qwen1.5B

Reinforcement Learning • 2B • Updated Jun 24, 2025 • 4

authored a paper about 1 year ago

Just Enough Thinking: Efficient Reasoning with Adaptive Length Penalties Reinforcement Learning

Paper • 2506.05256 • Published Jun 5, 2025 • 2

updated a collection about 1 year ago

Adaptive Length Penalty

Teaching language models to think efficiently with Adaptive Length Penalty (ALP) • 3 items • Updated Jun 24, 2025 • 1

authored 2 papers about 1 year ago

OpenThoughts: Data Recipes for Reasoning Models

Paper • 2506.04178 • Published Jun 4, 2025 • 56

The Common Pile v0.1: An 8TB Dataset of Public Domain and Openly Licensed Text

Paper • 2506.05209 • Published Jun 5, 2025 • 62

in SynthLabsAI/Big-Math-RL-Verified about 1 year ago

Release of data without filtering for solve rate

#3 opened over 1 year ago by

Solution of the Problems

#6 opened about 1 year ago by

authored a paper about 1 year ago

Learning Adaptive Parallel Reasoning with Language Models

Paper • 2504.15466 • Published Apr 21, 2025 • 44

updated a collection about 1 year ago

Big-Math

This collection contains assets associated with the Big-Math dataset, a high-quality collection of over 250,000 math questions with verifiable answers • 4 items • Updated Apr 16, 2025 • 7

in SynthLabsAI/Big-Math-RL-Verified over 1 year ago

Adding an indicator of whether response requires LLM judge

#4 opened over 1 year ago by