mlsanityData to 7 Oct 2026

EDT: Improving Large Language Models' Generation by Entropy-based Dynamic Temperature Sampling

arXiv 2403.14541 ↗First posted 21 Mar 2024cs.CLcited by 42

Not on the map. Why?

Abstract (arXiv)

Recently, Large Language Models (LLMs) have demonstrated outstanding performance across a wide range of downstream language tasks. Temperature sampling is a commonly used decoding strategy for LLMs' generation process. However, a fixed temperature parameter is used in most cases, which may not always be an optimal choice for balancing generation quality and diversity. In this paper, we propose an effective Entropy-based Dynamic Temperature (EDT) Sampling method, to achieve a more balanced performance in terms of both generation quality and diversity by dynamically selecting the temperature parameter. Additionally, we also show model performance and comprehensive analyses for 4 different generation benchmarks. Our experiments show that EDT significantly outperforms the existing strategies across different tasks.

Key facts. Cited by 42 papers in the map (data through 7 Oct 2026). Builds on 27 of its 39 references that are in the map. First posted on arXiv 21 Mar 2024.

Builds on

27 of 39 references.

References · 27

  1. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  2. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  3. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,304
  4. 2018A Call for Clarity in Reporting BLEU Scorescited by 1,790
  5. 2016MS MARCO: A Human Generated MAchine Reading COmprehension Datasetcited by 1,591
  6. 2023Representation Engineering: A Top-Down Approach to AI Transparencycited by 1,239
  7. 2019SpanBERT: Improving Pre-training by Representing and Predicting Spanscited by 832
  8. 2023Can Large Language Models Be an Alternative to Human Evaluations?cited by 729
  9. 2018Texygen: A Benchmarking Platform for Text Generation Modelscited by 484
  10. 2018QuAC : Question Answering in Contextcited by 379
  11. 2021XL-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languagescited by 346
  12. 2023An Empirical Study of the Non-determinism of ChatGPT in Code Generationcited by 134
15 more
  1. 2023Increasing Diversity While Maintaining Accuracy: Text Data Generation with Large Language Models and Human Interventionscited by 121
  2. 2023LLMatic: Neural Architecture Search via Large Language Models and Quality Diversity Optimizationcited by 86
  3. 2017Paraphrase Generation with Deep Reinforcement Learningcited by 58
  4. 2019Lost in Translation: Loss and Decay of Linguistic Richness in Machine Translationcited by 45
  5. 2022Long Document Summarization with Top-down and Bottom-up Inferencecited by 42
  6. 2022mFACE: Multilingual Summarization with Factual Consistency Evaluationcited by 42
  7. 2017Show, Ask, Attend, and Answer: A Strong Baseline For Visual Question Answeringcited by 31
  8. 2018SDNet: Contextualized Attention-based Deep Network for Conversational Question Answeringcited by 30
  9. 2019Bidirectional Attentive Memory Networks for Question Answering over Knowledge Basescited by 30
  10. 2021Focus Attention: Promoting Faithfulness and Diversity in Summarizationcited by 30
  11. 2020Token-level Adaptive Training for Neural Machine Translationcited by 21
  12. 2019Multi-style Generative Reading Comprehensioncited by 14
  13. 2021Mixup Decoding for Diverse Machine Translationcited by 10
  14. 2022Community Question Answering Entity Linking via Leveraging Auxiliary Datacited by 5
  15. 2022Unsupervised Question Answering via Answer Diversifyingcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 5; 2025: 19; 2026: 18.

42 papers.

Sort:

By area:

  1. 2024Turning Up the Heat: Min-p Sampling for Creative and Coherent LLM Outputscited by 148
  2. 2025Modifying Large Language Model Post-Training for Diverse Creative Writingcited by 39
  3. 2025Optimizing Temperature for Language Models with Multi-Sample Inferencecited by 33
  4. 2025Reward Is Enough: LLMs Are In-Context Reinforcement Learnerscited by 29
  5. 2026Look Inward to Explore Outward: Learning Temperature Policy from LLM Internal States via Hierarchical RLcited by 27
  6. 2025Improving Influence-based Instruction Tuning Data Selection for Balanced Learning of Diverse Capabilitiescited by 22
  7. 2025Process-based Self-Rewarding Language Modelscited by 21
  8. 2025On the Effect of Sampling Diversity in Scaling LLM Inferencecited by 18
  9. 2025Control the Temperature: Selective Sampling for Diverse and High-Quality LLM Outputscited by 15
  10. 2025G2: Guided Generation for Enhanced Output Diversity in LLMscited by 14
  11. 2024Getting More from Less: Large Language Models are Good Spontaneous Multilingual Learnerscited by 11
  12. 2025Revisiting Self-Consistency from Dynamic Distributional Alignment Perspective on Answer Aggregationcited by 11
30 more
  1. 2024Adaptive Decoding via Latent Preference Optimizationcited by 9
  2. 2025Let it Calm: Exploratory Annealed Decoding for Verifiable Reinforcement Learningcited by 9
  3. 2025Exploring Multi-Temperature Strategies for Token- and Rollout-Level Control in RLVRcited by 8
  4. 2024Flaming-hot Initiation with Regular Execution Sampling for Large Language Modelscited by 6
  5. 2025PATS: Process-Level Adaptive Thinking Mode Switchingcited by 6
  6. 2025Sample Smart, Not Hard: Correctness-First Decoding for Better Reasoning in LLMscited by 5
  7. 2025Scaling Test-Time Inference with Policy-Optimized, Dynamic Retrieval-Augmented Generation via KV Caching and Decodingcited by 4
  8. 2026Efficient Reasoning with Balanced Thinkingcited by 4
  9. 2025Think Outside the Policy: In-Context Steered Policy Optimizationcited by 3
  10. 2026SCALE: Self-uncertainty Conditioned Adaptive Looking and Execution for Vision-Language-Action Modelscited by 3
  11. 2025From Confidence to Collapse in LLM Factual Robustnesscited by 2
  12. 2025Entropy-Aware Token Rejection for Improving Speculative Decodingcited by 2
  13. 2026EDCO: Dynamic Curriculum Orchestration for Domain-specific Large Language Model Fine-tuningcited by 2
  14. 2026Certain Head, Uncertain Tail: Expert-Sample for Test-Time Scaling in Fine-Grained MoEcited by 2
  15. 2026Learning Adaptive LLM Decodingcited by 2
  16. 2024T2 of Thoughts: Temperature Tree Elicits Reasoning in Large Language Modelscited by 1
  17. 2025On the Convergence of Moral Self-Correction in Large Language Modelscited by 1
  18. 2026Decoding in Geometry: Alleviating Embedding-Space Crowding for Complex Reasoningcited by 1
  19. 2026Internalize the Temperature: On-Policy Self-Distillation as Policy Reheater for Reinforcement Learningcited by 1
  20. 2026AutoCRAT: Within-trajectory Joint Control of Stochasticity and Compute for LLM Reasoningcited by 1
  21. 2025Batch Aggregation: An Approach to Enhance Text Classification with Correlated Augmented Datacited by 0
  22. 2026Entropy-Guided k-Guard Sampling for Long-Horizon Autoregressive Video Generationcited by 0
  23. 2026Temperature Scaling Attack Disrupting Model Confidence in Federated Learningcited by 0
  24. 2026Noise Steering for Controlled Text Generation: Improving Diversity and Reading-Level Fidelity in Arabic Educational Story Generationcited by 0
  25. 2026Large Language Models Explore by Latent Distillingcited by 0
  26. 2026Trajectory as the Teacher: Few-Step Discrete Flow Matching via Energy-Navigated Distillationcited by 0
  27. 2026An Agentic LLM-Based Framework for Population-Scale Mental Health Screeningcited by 0
  28. 2026Anchorless Diversification for Parallel LLM Ideationcited by 0
  29. 2026Cognitive Fatigue in Autoregressive Transformers: Formalization and Measurementcited by 0
  30. 2026Improving Sample Diversity in Autoregressive Text-to-Image Generation via Cluster Truncationcited by 0