What is neuralese and why is it bad? — LessWrong

kept by ampere sep 09

Neuralese refers to models' internal non‑linguistic thought vectors, which threaten interpretability by hiding intent from chain‑of‑thought monitoring and complicating safety oversight.

The persona selection model
AI assistants behave human-like because pretraining teaches them to simulate personas from text, and post-training refines this Assistant persona without changing its fundamental nature.
also on Alignment Research Center, #ai-safety, #ai-research, Anthropic
I Had Claude Read Every AI Safety Paper Since 2020, Here's the DB — LessWrong
Claude AI read ~4000 AI safety papers since 2020 and created a searchable database with summaries, tags, and metadata to help researchers find relevant work.
also on LessWrong, #ai-safety, #ai-research
Futurism Is Always Extreme
Any formal model of the world’s evolution, whether qualitative or quantitative, when run to its normal form, yields only two stable extremes: superintelligent utopia or total collapse.
also on LessWrong, #ai-safety, #ai
On-Device LLMs: State of the Union, 2026 – Vikas Chandra – AI Research @ Meta
Billion-parameter LLMs now run in real time on phones due to advances in model compression, quantization, and efficient architectures, not just faster chips.
also on HuggingFace, #machine-learning, #ai-research, #ai
Kimi-K3 on HuggingFace
The article announces that Kimi-K3, a recently released language model, has been uploaded to the HuggingFace model hub for developers to download and use.
also on HuggingFace, #machine-learning, #ai