Starstrek

Stars321123's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

upvoted 7 papers 13 days ago

LatentPress: Context Compression Beyond Text and Vision

Paper • 2609.01507 • Published 18 days ago • 118

Repo-To-Skill: Distilling GitHub Repositories Into AI4AI Skills

Paper • 2609.02749 • Published 17 days ago • 399

Random Attention: Rethinking KV Cache Eviction for Efficient Reasoning

Paper • 2609.03430 • Published 16 days ago • 186

RoboTok: An Internet-Scale Data Engine for Human Demonstration Retrieval and Dexterous Manipulation Learning

Paper • 2609.03199 • Published 17 days ago • 31

LLaDA-Image: Building Strong Image Generators with Fully Open Training Recipes

Paper • 2609.03796 • Published 16 days ago • 183

Terminal-Universe: Turning Agent Trajectories into Scalable Terminal Environments

Paper • 2609.04148 • Published 16 days ago • 239

Compile by Training: Turning Natural-Language Specifications into Local Neural Functions

Paper • 2609.04199 • Published 16 days ago • 324

upvoted 2 collections 13 days ago

Kairos3.1

Collection

3 items • Updated Jul 8 • 2

ACE-Data

Collection

1 item • Updated 29 days ago • 2

upvoted 2 papers 15 days ago

S3Gym: Can LLMs Turn Self-Testing and Self-Judging into Self-Improvement?

Paper • 2608.31100 • Published 19 days ago • 40

Influence-Directed Distillation: Solving the Diversity Bottleneck in Sampled-Token On-Policy Distillation

Paper • 2608.29846 • Published 20 days ago • 15

upvoted a paper 16 days ago

HarnessDev: Can LLMs Create and Evolve Their Own Agent Harness?

Paper • 2609.01437 • Published 18 days ago • 117

upvoted 2 articles 16 days ago

view article

Article

Fine-tuning a 350M Model for Better Structured Outputs in 100 GRPO Steps

• 
• 
• +1

iamleonie, burtenshaw, sergiopaniego

•

16 days ago

• 124

view article

Article

Training a coding model to paint watercolours with TRL and OpenEnv

• 

sergiopaniego

•

16 days ago

• 68

upvoted 4 collections 16 days ago

MegaMath

Collection

MegaMath, the largest open math pre-training dataset curated from diverse, math-focused sources, with over 300B tokens. • 4 items • Updated 15 days ago • 5

K2-V2

Collection

The collection for K2-V2 models. • 7 items • Updated 15 days ago • 23

K2 Horizon

Collection

K2 Horizon models, datasets, and supporting resources • 22 items • Updated 8 days ago • 129

Spark-X2.5

Collection

Spark-X2.5 is a compact, general-purpose language model for conversation, writing, translation, reasoning, coding, tool use, and agentic workflows. • 10 items • Updated 3 days ago • 43

upvoted 2 papers 17 days ago

SMELT: Scaling Laws for Compute-Matched MoE Looped Transformers

Paper • 2609.01343 • Published 18 days ago • 89

On the Design of Qwen3.8-Next Architecture: Evaluation, Efficiency, and Training Stability

Paper • 2608.30320 • Published 19 days ago • 59