view article Article The 1 Billion Token Challenge: Finding the Perfect Pre-training Mix 21 days ago • 42
Mem-Agent Collection Small sized agents from Dria trained on interacting with an obsidian-like memory system using python tools. Trained on Qwen3-4B-Thinking-2507. • 4 items • Updated Sep 5 • 3
BeyondWeb: Lessons from Scaling Synthetic Data for Trillion-scale Pretraining Paper • 2508.10975 • Published Aug 14 • 60
view article Article mem-agent: Persistent, Human Readable Memory Agent Trained with Online RL Sep 11 • 25
view article Article Building Enterprise-Ready Text Classifiers in Minutes with Adaptive Learning Aug 9 • 12
On the Generalization of SFT: A Reinforcement Learning Perspective with Reward Rectification Paper • 2508.05629 • Published Aug 7 • 178
view article Article Unsupervised Model Improvement via Internal Coherence Maximization: Outperforming Human-Supervised Methods Through Self-Elicitation Aug 3 • 7
GLM-4.5 Collection GLM-4.5: An open-source large language model designed for intelligent agents by Z.ai • 11 items • Updated Aug 11 • 247
view article Article Understanding Model Reasoning Through Thought Anchors: A Comparative Study of Qwen3 and DeepSeek-R1 Jul 23 • 4
Ellora Collection Ellora: Enhancing LLMs with LoRA - Standardized Recipes for Capability Enhancement • 12 items • Updated Oct 20 • 4
Internal Coherence Maximization Collection Internal Coherence Maximization (ICM): A Label-Free, Unsupervised Training Framework for LLMs • 7 items • Updated Oct 10 • 2
Pre-training Dataset Samples Collection A collection of pre-training datasets samples of sizes 10M, 100M and 1B tokens. Ideal for use in quick experimentation and ablations. • 19 items • Updated 13 days ago • 13