ML Newsnew | past | best | rss | submit about
Stories from February 12, 2026 (UTC)
Go back a day, month, or year. Go forward a day.
1. NanoQuant: Efficient Sub-1-Bit Quantization of Large Language Models (arxiv.org)
13 points by chrsw 234 days ago | hide | past | pdf | discuss
2. Grok4 sabotages shutdown 97% of the time,even if instructed not in system prompt (arxiv.org)
8 points by agenticagent 234 days ago | hide | past | pdf | 4 comments
3. HySparse: A Hybrid Sparse Attention Architecture (arxiv.org)
5 points by readitalready 234 days ago | hide | past | pdf | discuss
4. RL on GPT-5 to write better kernels (arxiv.org)
4 points by atallahw 234 days ago | hide | past | pdf | 1 comment
5. Biases in the Blind Spot: Detecting What LLMs Fail to Mention (arxiv.org)
4 points by typeofhuman 234 days ago | hide | past | pdf | 1 comment
6. SoftMatcha 2: A Fast and Soft Pattern Matcher for Trillion-Scale Corpora (arxiv.org)
3 points by salkahfi 234 days ago | hide | past | pdf | discuss
7. Pushing Tensor Accelerators Beyond MatMul in a User-Schedulable Language (arxiv.org)
2 points by matt_d 234 days ago | hide | past | pdf | discuss
8. Opus: Towards Efficient and Principled Data Selection in LLM Pre-Training (arxiv.org)
2 points by onurkanbkrc 234 days ago | hide | past | pdf | discuss
9. Faster and Cheaper Computations with Randomized Numerical Linear Algebra (arxiv.org)
2 points by PaulHoule 234 days ago | hide | past | pdf | discuss
10. Routing LLM queries using internal success predictions (70% cost reduction) (arxiv.org)
1 point by stansApprentice 234 days ago | hide | past | pdf | 3 comments
11. SWE-AGI: benchmarking spec-driven software construction (arxiv.org)
1 point by mustaphah 234 days ago | hide | past | pdf | 1 comment
12. Biases in the Blind Spot: Detecting What LLMs Fail to Mention (arxiv.org)
1 point by jari_mustonen 234 days ago | hide | past | pdf | discuss
13. Evaluation of RAG Architectures for Policy Document Question Answering (arxiv.org)
1 point by PaulHoule 234 days ago | hide | past | pdf | discuss
14. Step 3.5 Flash: Open Frontier-Level Intelligence with 11B Active Parameters (arxiv.org)
1 point by onurkanbkrc 234 days ago | hide | past | pdf | 1 comment
About | RSS | RSS (all) | HN arXiv