Metraipolis › archive › dc_0ccd5b60ffe0

GlitchPatch: Repairing Glitch Tokens in Frozen Language Models via Local Retokenization — arXiv:2610.04399

Source: arXiv
Original: https://arxiv.org/abs/2610.04399
License (as archived): CC0-1.0
Archived: 2026-10-06T05:26:11.249Z
SHA-256 of the archived text: 589df92117f972a7ac543aa0e84963bc2b1fe197e1e92cdefd7e0da2019f08b8
Stable URL of this page: https://metraipolis.com/documento/dc_0ccd5b60ffe0

How to cite

GlitchPatch: Repairing Glitch Tokens in Frozen Language Models via Local Retokenization — arXiv:2610.04399 — arXiv, https://arxiv.org/abs/2610.04399 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0ccd5b60ffe0 (Metraipolis archive dc_0ccd5b60ffe0, sha256 589df92117f972a7ac543aa0e84963bc2b1fe197e1e92cdefd7e0da2019f08b8)

BibTeX

@misc{dc0ccd5b60ffe0,
  title = {GlitchPatch: Repairing Glitch Tokens in Frozen Language Models via Local Retokenization — arXiv:2610.04399},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.04399},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0ccd5b60ffe0; sha256 589df92117f972a7ac543aa0e84963bc2b1fe197e1e92cdefd7e0da2019f08b8},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.04399. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Glitch tokens are anomalous vocabulary entries that can cause large language models (LLMs) to produce outputs inconsistent with their inputs. Existing repair methods require access to model internals, making them impractical for frozen checkpoints. We investigate whether glitch tokens can be repaired outside the model by optimizing the input tokenization. An empirical study on BPE merge-rule deletion reveals that (1)deleting a glitch token's merge rule can fix a substantial fraction of failures, yet disrupting normal tokens sharing intermediate merge nodes causes the overall glitch rate to rise, and (2)different decomposition granularities yield non-monotonic fix rates while collateral damage on normal tokens grows monotonically. Motivated by these findings, we propose GlitchPatch, a repair framework for frozen language models based on local retokenization, consisting of two stages: the offline stage uses Behavioral Path Optimization (BPO) to find the behaviorally optimal replacement token sequence for each glitch token and compiles validated replacements into a rule table; the online stage substitutes only the IDs of matched glitch tokens in the canonical token sequence, with no modification to model parameters or internal states. Experiments on ten models spanning six tokenizer families show that GlitchPatch achieves an 85.10% mean fix rate, outperforming the strongest baseline by 14.37 percentage points, and reduces the average glitch rate from 14.88% to 2.27%. GlitchPatch achieves a 0.00% RR in full-vocabulary evaluation and leaves rule-unmatched inputs unchanged by design. We further evaluate the practical impact of repair from the perspectives of time cost, language understanding, and capability, supporting its deployment feasibility. We hope this work provides a practical option for improving tokenizer reliability.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.04399 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-06T05:24:05.844Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).