Skip to content
AI News HubLIVE
Original source3 min read

Can an LLM dream? Testing the overfitted brain hypothesis on an agent with text memory

Summary

A study tests the overfitted brain hypothesis on a text-memory LLM agent: across 10 simulated days and 30 runs, corrupting conversation transcripts did not improve transfer, 40% detail corruption lowered recall from 76% to 60%, and random-text corruption had no consistent effect because the model recognized and excluded it.

SourcearXiv Computational LinguisticsAuthor: Eric Zou
Can an LLM dream? Testing the overfitted brain hypothesis on an agent with text memory
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[Submitted on 14 Oct 2026]

Title:Can an LLM dream? Testing the overfitted brain hypothesis on an agent with text memory

View a PDF of the paper titled Can an LLM dream? Testing the overfitted brain hypothesis on an agent with text memory, by Eric Zou

View PDF HTML (experimental)

Abstract:Hoel (2021) proposed that dreams distort replayed experience so that the brain does not overfit to recent events. We test this hypothesis on an LLM agent (Claude Opus 5.5) that stores what it learns in a text memory and rewrites that memory each night from the day's conversation transcript. The agent goes through ten simulated days of conversation built from LongMemEval. We corrupt 5, 10, 20 or 40 percent of the transcript's turns before each nightly rewrite in one of two ways. Detail corruption changes one or two details in a turn to plausible wrong ones, or replaces the turn with one by the same speaker from an earlier day. Random-text corruption replaces half of the selected turns with a passage of Moby-Dick and inserts Moby-Dick words into the other half. A size limit on the memory, rising from 19,500 to 117,000 tokens over ten nights, set from human free-recall data, forces it to forget. We learned that corruption did not help the agent apply what it learned in earlier conversations to new tasks in any of 30 runs (ten conditions, three different conversation histories). Replacing details in 40 percent of turns lowered recall from 76 to 60 percent. Random text had no consistent effect on recall, with a paired difference of 0 $\pm$ 2 points at 40 percent, because the model recognized it as corrupted and left it out of memory. Replaced details were stored as fact even when another turn held the correct value, and the agent answered 40 percent of those questions correctly, compared with 75 percent without corruption. These results suggest that a pretrained model can separate corrupted text from real events, so a fair test needs a learner that cannot make this separation as easily, such as a model fine-tuned on the corrupted days or an agent whose days take place in an invented world.

Comments: 13 pages, 7 figures, 2 tables. Code and data: this https URL

Subjects:

Computation and Language (cs.CL); Artificial Intelligence (cs.AI)

Cite as: arXiv:2611.21648 [cs.CL]

(or arXiv:2611.21648v1 [cs.CL] for this version)

https://doi.org/10.48550/arXiv.2611.21648

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Eric Jujian Zou [view email] [v1] Wed, 14 Oct 2026 20:44:10 UTC (104 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled Can an LLM dream? Testing the overfitted brain hypothesis on an agent with text memory, by Eric Zou

View PDF

HTML (experimental)

TeX Source

view license

Additional Features

Audio Summary

Current browse context:

cs.CL

new | recent | 2026-11

Change to browse by:

cs cs.AI

References & Citations

SciX

Google Scholar

Semantic Scholar

Loading...

Data provided by:

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 for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • Across 30 runs, transcript corruption at any tested rate did not help the agent transfer earlier learning to new tasks.
  • Replacing details in 40% of turns lowered recall from 76% to 60%.
  • Random-text corruption had no consistent recall effect, with a paired difference of 0 ± 2 points at 40%.
  • The model excluded obvious foreign text but stored replaced details as fact, even when another turn held the correct value.

Highlights and analysis are generated automatically and may contain errors. Check the original source.