[2608.18578] Compress and Forget: bitsandbytes Quantization Amplifies Proactive Interference in LLMs

Skip to main content

Search arXiv

Press Enter to search · Advanced search

-->

Computer Science > Computation and Language

arXiv:2608.18578 (cs)

[Submitted on 19 Aug 2026 (v1), last revised 20 Aug 2026 (this version, v2)]

Title:Compress and Forget: bitsandbytes Quantization Amplifies Proactive Interference in LLMs

Authors:Shayan Shahrabi-Farahani, Dara Rahmati<br>View a PDF of the paper titled Compress and Forget: bitsandbytes Quantization Amplifies Proactive Interference in LLMs, by Shayan Shahrabi-Farahani and 1 other authors

View PDF<br>HTML (experimental)

Abstract:Proactive interference (PI) is a documented failure mode in large language models in which retrieval of a repeatedly overwritten value degrades as prior overwrites accumulate, mirroring a classical phenomenon in human working memory. Post-training quantization (PTQ) is now the default deployment path for open-weight models, yet its effect on this failure mode has not been tested. We evaluate three precision levels (FP16, INT8, INT4/NF4, via bitsandbytes) across three architecturally distinct instruction-tuned models (Qwen2.5-7B-Instruct, Mistral-7B-Instruct-v0.3, Phi-3.5-mini-instruct), holding the retrieval task fixed. INT4 quantization significantly reduces accuracy under high interference in every model (e.g., from 81.0% to 68.3% for Qwen), confirmed by paired McNemar's tests ($p \le 2.6 \times 10^{-6}$) and a mixed-effects regression spanning all interference levels; INT8, often assumed safe, also carries a smaller but real penalty in two of three models. The effect is specific to semantically similar (word-type) distractors and reverses sign under a numeric control condition, and is mechanistically linked to a rise in same-key intrusion errors under INT4 (from 21.5% to 24.6% of trials, $p = 4.8 \times 10^{-7}$). A follow-up ablation shows the effect originates in the quantized transformer backbone rather than the output projection layer. These results suggest that bitsandbytes 4-bit quantization can impose an additional cost on applications relying on long, updatable, semantically dense contexts, even when aggregate benchmark accuracy appears largely unaffected. We release our code and tokenizer-verified vocabulary construction method at this https URL

Comments:<br>21 pages, 6 figures, 11 tables. Author list formatting simplified. Code and data released at this https URL

Subjects:

Computation and Language (cs.CL); Machine Learning (cs.LG)

ACM classes:<br>I.2.7; I.2.6

Cite as:<br>arXiv:2608.18578 [cs.CL]

(or<br>arXiv:2608.18578v2 [cs.CL] for this version)

https://doi.org/10.48550/arXiv.2608.18578

Focus to learn more

arXiv-issued DOI via DataCite

Submission history<br>From: Shayan Shahrabi-Farahani [view email]<br>[v1]<br>Wed, 19 Aug 2026 06:17:13 UTC (347 KB)

[v2]<br>Thu, 20 Aug 2026 09:45:51 UTC (347 KB)

Full-text links:<br>Access Paper:

View a PDF of the paper titled Compress and Forget: bitsandbytes Quantization Amplifies Proactive Interference in LLMs, by Shayan Shahrabi-Farahani and 1 other authors<br>View PDF<br>HTML (experimental)<br>TeX Source

view license

Current browse context:

cs.CL

next >

new<br>recent<br>| 2026-08

Change to browse by:

cs<br>cs.LG

References & Citations

NASA ADS<br>Google Scholar

Semantic Scholar

export BibTeX citation<br>Loading...

BibTeX formatted citation

&times;

loading...

Data provided by:

Bookmark

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea...