GPT-2 XL

Researchers introduced TreeTracer, a visual analytics tool that aggregates hundreds of stochastic LLM generations into syntax-aligned trees to expose hidden biases. The tool uses perturbation analysis…

Researchers at arXiv (arXiv:2605.23035) used sparse autoencoders to decompose GPT-2 XL and Llama-3.1-8B into 16K-32K interpretable features per layer, finding that semantic features alone recovered 94…

Researchers decomposed GPT-2 XL and Llama-3.1-8B into 16,000 to 32,000 interpretable features per layer using sparse autoencoders, finding that semantic features alone recovered 94% of peak brain-resp…