Pith. sign in

REVIEW 1 cited by

Contextual Temperature for Language Modeling

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2012.13575 v1 pith:YYY7CINZ submitted 2020-12-25 cs.CL cs.LG

Contextual Temperature for Language Modeling

classification cs.CL cs.LG
keywords temperatureoptimalschedulesapproachcontextcontextualfixedlanguage
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Temperature scaling has been widely used as an effective approach to control the smoothness of a distribution, which helps the model performance in various tasks. Current practices to apply temperature scaling assume either a fixed, or a manually-crafted dynamically changing schedule. However, our studies indicate that the individual optimal trajectory for each class can change with the context. To this end, we propose contextual temperature, a generalized approach that learns an optimal temperature trajectory for each vocabulary over the context. Experimental results confirm that the proposed method significantly improves state-of-the-art language models, achieving a perplexity of 55.31 and 62.89 on the test set of Penn Treebank and WikiText-2, respectively. In-depth analyses show that the behaviour of the learned temperature schedules varies dramatically by vocabulary, and that the optimal schedules help in controlling the uncertainties. These evidences further justify the need for the proposed method and its advantages over fixed temperature schedules.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Sampling for Quality: Training-Free Reward-Guided LLM Decoding via Sequential Monte Carlo

    cs.LG 2026-04 unverdicted novelty 7.0

    Sequential Monte Carlo sampling from a reward-augmented sequence distribution improves LLM performance on HumanEval by up to 54.9% and MATH500 by up to 8.8%, outperforming standard sampling and GRPO.