A token-level analysis of this failure mode is presented by viewing decoding as a dynamical process that enters and persists in a small set of recurrent contexts and shows that persistence is controlled by the escape mass assigned to plausible alternatives within the token sampling set.
Abstract
Pruning is a practical approach to compress large language models (LLMs), but it can amplify text degeneration, especially repetition loops, even when perplexity and task accuracy remain largely unchanged. In this work, we present a token-level analysis of this failure mode by viewing decoding as a dynamical process that enters and persists in a small set of recurrent contexts. Our analysis decomposes degeneration into loop entry risk and loop persistence, and shows that persistence is controlled by the escape mass assigned to plausible alternatives within the token sampling set. Motivated by these findings, we propose two token-level guidance objectives for post-pruning fine-tuning. FOCUS reweights distillation toward high-confidence teacher regions to suppress leakage, while RePAIR uses onset-centered positive/negative continuation pairs with a margin loss to promote plausible alternatives and prevent early commitment to repetition loops. Experiments on open-ended continuation and instruction-based generation show that both methods consistently reduce repetition and improve generation quality.
Attention Diversion explains why inference-time strategies, including chain-of-thought prompting, spell-checking, self-repair, and stronger repair models, fail to consistently recover performance: each addresses one channel at a time.
Jiaqi Zhu, Yang Zhang, Junhua Ding et al.· 0 citations
ADU is presented, a fine-grained, training-based framework that shifts unlearning from token erasure to contextual attention-pathway decoupling, and achieves the strongest aggregate performance among evaluated baselines on the TOFU and WMDP benchmarks.
Xunlei Chen, Qirui Ye, Yuang Li et al.· 0 citations
This work instantiates budgeted oracle-to-hint compression in online judge (OJ) style algorithmic programming as a modular interactive agent that couples an LLM core with a sandboxed judger, a feedback-to-hint prompt constructor, and trajectory memory.
Jialiang Gu, Keren Zhou, Daming Li et al.· SIGSOFT FSE Companion· 2 citations
Visual token pruning reduces the inference cost of multimodal large language models, but a fixed token ratio is poorly matched to text-rich inputs. In OCR-centric tasks, decisive evidence can be a small number, label, or field whose relevance is specified by the question; indiscriminate pruning can erase that evidence while retaining visually salient but irrelevant regions. We present ET-Prune, a training-free framework that casts pruning as evidence allocation. It derives question-conditioned evidence from a decoder-side partial query-key block, safeguards text-like spatial regions, and converts evidence uncertainty and density into a sample-specific token floor. Three progressive middle-layer events then move the sequence toward this budget, retaining more tokens for diffuse or text-dense evidence and pruning concentrated evidence more aggressively. At the observed point estimates from one deterministic pass per configuration, ET-Prune leads or ties among pruned methods in all six backbone-benchmark comparisons at roughly half tokens. On OCRBench-v2, it leads the strongest pruned baselines by 1.80 and 0.68 percentage points on Qwen3-VL-8B and InternVL3.5-8B, respectively, while retaining about half of the visual tokens; on MMBench v1.1, it reaches 0.8467 circular exact-matching accuracy versus 0.8437 for Vanilla at 54.45% average visual-token retention. These results show a favorable observed quality-cost trade-off for evidence-aware dynamic budgeting in text-rich multimodal inference.
Zizhong Ding, Junxian Li, Kai Liu et al.· 0 citations
It is concluded that mean attention degradation is largely descriptive rather than prescriptive: function tokens contribute through what their hidden states compute, not through the attention they receive -- with implications for interpretability methodology and attention-score-based inference optimisations such as KV-cache eviction.
This work finds that models cannot reliably reconstruct the original meaning and layer-wise probe classifiers fail to detect the harmful intent of perturbed prompts, and perturbations can occasionally reduce attack success by inducing off-topic or incoherent responses.