codelion

3,064 karmajoined February 23, 2008On HN

Recent submissions

Cross-Lingual Alignment Without Joint Training arxiv.org
1codelion1ddiscuss
Voice AI Benchmarks: a directory of voice-agent evaluations voicebenchmarks.com
2codelion2ddiscuss
Neural Canvas: a fruit fly connectome demo in the browser huggingface.co
2codelion4ddiscuss
Iris: Climbing to the Search Frontier arxiv.org
1codelion4ddiscuss
The Recursive Economy: AI Self-Improvement and Scarcity recursive-economy.pages.dev
2codelion24ddiscuss
Neural Drive, a SuperTuxKart World Model that runs in the browser huggingface.co
2codelion1modiscuss
Coding agent for local models on a Mac mlx-optiq.com
1codelion1modiscuss
Frame: Grounding LLM Vulnerability Detection with a Sound Separation-Logic Core lambdasec.github.io
2codelion2modiscuss
Mlx-optiq: per-layer mixed-precision LLM quantization for Apple Silicon mlx-optiq.com
2codelion2modiscuss
A 1B humanizer that matches human writing on an AI detector mlx-optiq.com
3codelion3modiscuss
Zero-Dependency Programming conjure.pages.dev
2codelion5modiscuss
Scaling Pedagogical Pre-Training: From Optimal Mixing to 10B Tokens huggingface.co
2codelion6modiscuss
From HashHop to Memory-Augmented Language Models huggingface.co
2codelion7modiscuss
The Optimal Architecture for Small Language Models huggingface.co
1codelion8modiscuss
Enhancing LLMs with LoRA – Standardized Recipes for Capability Enhancement huggingface.co
3codelion8modiscuss
OpenEvolve: Teaching LLMs to Discover Algorithms Through Evolution algorithmicsuperintelligence.ai
53codelion9mo9 comments
Ellora: Enhancing LLMs with LoRA Standardized Recipes for Capability Enhancement huggingface.co
1codelion9modiscuss
The 1B Token Challenge: Finding the Perfect Pre-Training Mix huggingface.co
6codelion9modiscuss
The 1B Token Challenge: Finding the Perfect Pre-Training Mix huggingface.co
3codelion10modiscuss
Pivotal Token Search (PTS): Targeting Critical Decision Points in LLM Training huggingface.co
2codelion1y1 comment
Internal Coherence Maximization(ICM): Label-Free Unsupervised Training Framework github.com/codelion
1codelion1ydiscuss
Unsupervised Model Improvement via Internal Coherence Maximization huggingface.co
1codelion1ydiscuss
Show HN: PTS Library – Analyze LLM reasoning through "thought anchors"
2codelion1ydiscuss
Automated Discovery of High-Performance GPU Kernels with OpenEvolve huggingface.co
4codelion1ydiscuss
Adaptive Classifier: Dynamic Text Classification with Continuous Learning huggingface.co
1codelion1ydiscuss
Show HN: DeepThink Plugin – Bring Gemini 2.5's parallel reasoning to open models
4codelion1ydiscuss
Eliciting Fine-Tuned Transformer Capabilities via Inference-Time Techniques arxiv.org
1codelion1ydiscuss
Show HN: System Prompt Learning – LLMs Learn Problem-Solving from Experience
48codelion1y13 comments

All submissions and comments on Hacker News.