Papers with token selection
RoBSA: RoPE-based Blockwise Sparse Multi-head Latent Attention (2026.acl-long)
Copied to clipboard
| Challenge: | Large Language Models (LLMs) have advanced in recent years, scaling up in both parameter count and context length. |
| Approach: | They propose a method to compute attention over a subset of context tokens and to implement token selection in a blockwise manner. |
| Outcome: | The proposed method reduces end-to-end inference latency by up to 2.55x with minimal accuracy loss compared to full attention in long-context scenarios for very large models. |
LAMB: A Training-Free Method to Enhance the Long-Context Understanding of SSMs via Attention-Guided Token Filtering (2025.acl-short)
Copied to clipboard
Zhifan Ye, Zheng Wang, Kejing Xia, Jihoon Hong, Leshu Li, Lexington Whalen, Cheng Wan, Yonggan Fu, Yingyan Celine Lin, Souvik Kundu
| Challenge: | Recent work attributes performance degradation to an exponential decay in hidden-state memory. |
| Approach: | They propose a token filtering strategy that is training-free and attention-guided . they propose 'LAMB' to preserve critical tokens during inference . |
| Outcome: | The proposed token filtering improves long-context performance by 30.35% over state-of-the-art methods on benchmarks. |
Diffusion Language Model Inference with Monte Carlo Tree Search (2026.findings-eacl)
Copied to clipboard
Zheng Huang, Kiran Ramnath, Yueyan Chen, Aosong Feng, Sangmin Woo, Balasubramaniam Srinivasan, Zhichao Xu, Kang Zhou, Shuai Wang, Haibo Ding, Lin Lee Cheong
| Challenge: | Existing methods for inference use heuristics to determine which positions to unmask and which tokens to commit . MEDAL is an inference-time scaling framework that integrates Monte Carlo Tree SEarch initialization for Diffusion Language Model inference. |
| Approach: | They propose a framework that integrates Monte Carlo Tree SEarch initialization for Diffusion Language Model inference. |
| Outcome: | The proposed framework achieves 22.0% improvement over existing inference strategies across multiple benchmarks. |
CA-GAR: Context-Aware Alignment of LLM Generation for Document Retrieval (2025.findings-acl)
Copied to clipboard
| Challenge: | Recent techniques such as Generation-Augmented Retrieval (GAR) and Generative Document Retrieleval (GDR) leverage LLMs to enhance retrieval performance but face key challenges: GAR’s generated content may not always align with the target document corpus, while GDR limits the generative capacity of LLM. |
| Approach: | They propose a Context-Aware Generation-Augmented Retrieval approach which integrates corpus information into their generation process. |
| Outcome: | Experimental results show that CA-GAR outperforms existing methods on seven tasks and four non-English languages. |
Invisible Entropy: Towards Safe and Efficient Low-Entropy LLM Watermarking (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods to watermark low-entropy content are expensive and risky . IE reduces parameter size by 99% while achieving performance on par with state-of-the-art methods . |
| Approach: | They propose a logit-based watermarking paradigm that uses entropy-based features to predict whether the next token is high or low. |
| Outcome: | The proposed method reduces parameter size by 99% while achieving performance on par with state-of-the-art methods. |
Selective Preference Optimization via Token-Level Reward Function Estimation (2025.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for maximizing preference optimization on all available tokens are noisy and inefficient. |
| Approach: | They propose a selective alignment strategy that centers on efficient key token selection without strong, fine-grained supervision signals. |
| Outcome: | The proposed strategy outperforms baseline methods on three benchmarks with up to 60% reduction in training hours. |
GraphKV: Breaking the Static Selection Paradigm with Graph-Based KV Cache Eviction (2025.emnlp-main)
Copied to clipboard
| Challenge: | Efficient Key-Value (KV) cache management is essential for processing long text sequences in large language models (LLMs). |
| Approach: | They propose a graph-based framework that redefines token selection for KV cache compression. |
| Outcome: | The proposed framework can be used in existing KV cache eviction methods such as SnapKV and PyramidKV in a plug-and-play manner. |
Uncertainty-Aware Contrastive Decoding (2025.findings-acl)
Copied to clipboard
| Challenge: | Large language models excel in a wide range of tasks, but generating factually accurate outputs remains a challenge. |
| Approach: | They propose a method that dynamically adjusts model contributions at each decoding step based on uncertainty. |
| Outcome: | The proposed method significantly improves factual accuracy and reliability over existing methods. |
SSG: Logit-Balanced Vocabulary Partitioning for LLM Watermarking (2026.acl-long)
Copied to clipboard
| Challenge: | Large language models can generate highquality, human-like content, but they also pose risks such as infringement of proprietary interests, misuse of outputs, and spread harmful misinformation. |
| Approach: | They propose a method that partitions the vocabulary into two logit-balanced subsets and lifts the lower bound of watermark strength for each token prediction. |
| Outcome: | The proposed method lifts the lower bound of watermark strength for each token prediction, thereby improving watermark detectability. |
Reward-Weighted Sampling: Enhancing Non-Autoregressive Characteristics in Masked Diffusion LLMs (2025.emnlp-main)
Copied to clipboard
| Challenge: | Masked diffusion models (MDMs) are promising non-autoregressive alternatives for large language modeling. |
| Approach: | They propose a method that leverages an external reward model to provide a principled global signal during the iterative diffusion process. |
| Outcome: | The proposed method improves non-autoregressive generation orders and performance across evaluation metrics. |