Challenge: Existing sampling methods that are sensitive to temperature scaling fail to distinguish between diversity and noise.
Approach: They propose a method that identifies informative tokens by eliminating noise directly in logit space and a new sampling method that is temperature-invariant.
Outcome: The proposed method outperforms existing methods with significant improvements in reasoning and creative writing tasks.

Similar Papers

Min-k Sampling: Decoupling Truncation from Temperature Scaling via Relative Logit Dynamics (2026.acl-long)

Copied to clipboard

Challenge: Existing methods for decoding large language models have extreme sensitivity to temperature parameter T.
Approach: They propose a dynamic truncation strategy that analyzes the local shape of the logit distribution to identify "semantic cliffs" they show that Min-k consistently improves text quality even under extreme temperature settings .
Outcome: The proposed method achieves strict temperature invariance and low sensitivity to hyperparameter choices.
What Tokens Truly Matter? The Logit Conflation Problem in LLM Sampling (2026.findings-acl)

Copied to clipboard

Challenge: Existing methods for large language models filter tokens based on logit magnitudes or derived statistics, under the implicit assumption that high-logit tokens are desirable.
Approach: They propose to isolate the Logit Conflation Problem by using attention-weighted attribution to extract prompt-relevance from token logits.
Outcome: The proposed method improves on LLaMA-3 and is training-free and low latency.
Balancing Diversity and Risk in LLM Sampling: How to Select Your Method and Parameter for Open-Ended Text Generation (2025.acl-long)

Copied to clipboard

Challenge: Recent studies suggest that sampling-based decoding strategies can be used to optimize the output of Large Language Models (LLMs) . previous studies have shown that likelihood-maximization produces degenerate text which contains repetitive loops and incoherent context, especially in open-ended tasks.
Approach: They propose to use a prefix tree to estimate the intrinsic capacity of a truncation sampling method by considering the trade-off between diversity and risk at each decoding step.
Outcome: The proposed method is based on a prefix tree which preserves the context of a full sentence.
Watermarking Large Language Models: An Unbiased and Low-risk Method (2025.acl-long)

Copied to clipboard

Challenge: Recent advances in large language models (LLMs) have highlighted the risk of misusing them, raising the need for accurate detection of LLM-generated content.
Approach: They propose a method to inject imperceptible identifiers into large language models (LLMs) this method is unbiased and preserves the original token distribution in expectation .
Outcome: The proposed method preserves the original token distribution in expectation and has lower risk of producing unsatisfactory outputs in low-entropy scenarios compared to existing unbiased watermarks.
Increasing Diversity While Maintaining Accuracy: Text Data Generation with Large Language Models and Human Interventions (2023.acl-long)

Copied to clipboard

Challenge: Large language models (LLMs) can be used to generate text data for training and evaluating other models.
Approach: They propose to use logit suppression and temperature sampling to diversify text generation but at the cost of data accuracy.
Outcome: The proposed approach can increase diversity but at the cost of data accuracy.
How Sampling Affects the Detectability of Machine-written texts: A Comprehensive Study (2025.findings-emnlp)

Copied to clipboard

Challenge: Recent detectors report near-perfect accuracy, often boasting AUROC scores above 99%, but these claims typically assume fixed generation settings, leaving open the question of how robust such systems are to changes in decoding strategies.
Approach: They examine how sampling-based decoding impacts detectability with a focus on how subtle variations in a model’s (sub)word-level distribution affect detection performance.
Outcome: The proposed framework systematically examines how sampling-based decoding impacts detectability, with a focus on how subtle variations in a model’s (sub)word-level distribution affect detection performance.
Let’s Sample Step by Step: Adaptive-Consistency for Efficient Reasoning and Coding with LLMs (2023.emnlp-main)

Copied to clipboard

Challenge: Existing methods for improving the correctness of output from large language models generate a constant number of samples per question, but Adaptive-Consistency reduces sample budget by up to 7.9 times with an average accuracy drop of less than 0.1%.
Approach: They propose a model-agnostic technique that dynamically adjusts the number of samples per question using a lightweight stopping criterion.
Outcome: The proposed technique reduces sample budget by 7.9 times with an average accuracy drop of less than 0.1%.
Large Language Models Are Bad Dice Players: LLMs Struggle to Generate Random Numbers from Statistical Distributions (2026.acl-long)

Copied to clipboard

Challenge: Existing large language models lack a functional internal sampler to faithfully sample from specified probability distributions . lack of robust sampling mechanisms across diverse application scenarios is a critical functional requirement .
Approach: They propose to use a dual-protocol design to disentangle failure modes . batch generation achieves only modest statistical validity, while independent requests collapse almost entirely .
Outcome: The proposed model fails to enforce uniform answer-position constraints and violates demographic targets in attribute-constrained text-to-image prompt synthesis.
Low-Perplexity LLM-Generated Sequences and Where To Find Them (2025.acl-srw)

Copied to clipboard

Challenge: Large Language Models (LLMs) are increasingly applied across various domains, but the ways they leverage their training data during inference remains only partially understood.
Approach: They propose a systematic approach that analyzes low-perplexity sequences and traces them back to their sources in the training data.
Outcome: The proposed pipeline extracts low-perplexity sequences across diverse topics while avoiding degeneration, then trace them back to their sources in the training data.
Found in the Middle: Permutation Self-Consistency Improves Listwise Ranking in Large Language Models (2024.naacl-long)

Copied to clipboard

Challenge: Large language models exhibit positional bias in how they use context, which affects listwise ranking.
Approach: They propose a method to marginalize out different list orders in the prompt to produce an order-independent ranking with less positional bias.
Outcome: The proposed method improves on five datasets in sorting and passage reranking by 34-52% . it marginalizes out different list orders in the prompt to produce an order-independent ranking .

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations