🎯 LoRA

The items that name it 9

• [AINews] Google admits AI Mode hurts Search revenue; -15% $500B drop worst since 2004 IPO. AI Bear Market coming?
  article2026-10-31

  Low-rank adapter fine-tuning; used for Mistral Large 4 with NVFP4 experts and merge-aware LoRA, and for the RL-trained harness adapter on LFM2.5-24B-A2B.

• Artificial Analysis (@ArtificialAnlys) on X
  thread2026-10-30

  Hyperflow H3 is an 8-step LoRA built on MiniMax H3, released with open weights.

• nemotron-halloween
  writeup2026-10-30

  Adapter trained on over 5000 reviewed multi-turn conversations to give Bolt E4B the grumpy pirate personality. Training took about one to two hours per model.

• A Fine-Grained Analysis of the LoRA Fine-Tuning Landscape with Implications for Data Selection
  paper2026-10-05It is written here as LoRA (Low-Rank Adaptation).

  Low-Rank Adaptation

• HeuFouFT: Task-Guided Metaheuristic Coordinate Search for Fourier Fine-Tuning
  paper2026-10-05
• LLM2Jev: LLMs Are Already Jev-Style Decision Models -- When and How to Fine-Tune Them
  paper2026-10-01
• EvE: An Alternate Optimizer to Adam
  paper2026-09-28It is written here as LoRA (fine-tuning).

  fine-tuning

• New LoRA Skills Should Read but Never Write
  paper2026-09-25It is written here as LoRA (low-rank adapters).

  low-rank adapters

• Fine-tuning a 350M Model for Better Structured Outputs in 100 GRPO Steps
  writeup2026-09-03It is written here as LoRA (Low-Rank Adaptation).

  Low-Rank Adaptation