conftrace_

logit scaling

3 papers

Explore in graph

Co-occurring keywords · papers mentioning both

large language model (2) entropy neuron (1) subword style (1) non-autoregressive generation (1) token sampling (1) n-gram model (1) inference time (1) layer normalization (1) token frequency (1) reward model (1)

Papers

Reward-Weighted Sampling: Enhancing Non-Autoregressive Characteristics in Masked Diffusion LLMs EMNLP 2025 Transferring Extreme Subword Style Using Ngram Model-Based Logit Scaling NAACL 2025 Confidence Regulation Neurons in Language Models NeurIPS 2024