The Salt - Curated AI
Subscribe
Sign in
Home
Notes
AI Notebooks
AI Repositories
Related Articles
deep dive
Archive
About
Latest
Top
Discussions
When Quantization Improves Reinforcement Learning
The Weekly Salt #91
7 hrs ago
•
Benjamin Marie
3
Could GRPO Be an "Off-Policy" Algorithm?
The Weekly Salt #90
Oct 9
•
Benjamin Marie
8
Pre-Training Updates: NVFP4 and Thinking Augmented
The Weekly Salt #89
Oct 1
•
Benjamin Marie
2
September 2025
How Poor SFT Data Overwrites Learned Knowledge
The Weekly Salt #87
Sep 24
•
Benjamin Marie
5
Jet-Nemotron: Searching for the Best Attention Architecture
DeltaNet + Hardware-aware Search
Sep 23
•
Benjamin Marie
2
MMBERT as a Drop-in Successor to XLM-R
The Weekly Salt #86
Sep 17
•
Benjamin Marie
1
LLMs Hallucinate and That's a Benchmarking Problem
The Weekly Salt #85
Sep 10
•
Benjamin Marie
4
What Breaks When You Quantize for Translation? A Deep Dive Across 55 Languages
Evaluating LLM translation under quantization with COMET, BLEU, GGUF models, and more
Sep 8
•
Benjamin Marie
1
2
Joint Prediction of Token Order and Next Token
The Weekly Salt #84
Sep 3
•
Benjamin Marie
2
August 2025
Dual Preference Optimization with DuPO
The Weekly Salt #83
Aug 27
•
Benjamin Marie
3
Easier MoE Training and GFPO, a New Alternative to GRPO Against "Length Inflation"
The Weekly Salt #82
Aug 20
•
Benjamin Marie
2
Better MoE Compression and Understanding of Massive Activations
The Weekly Salt #81
Aug 13
•
Benjamin Marie
5
This site requires JavaScript to run correctly. Please
turn on JavaScript
or unblock scripts
