Paper Detail

WMIP: A Working-Memory-Inspired Programmable Probe of Cumulative State Tracking in Large Language Models

Dengzhe Hou, Lingyu Jiang, Deng Li, Zirui Li, Fangzhou Lin, Kazunori D Yamada

huggingface Score 9.0

Published 2026-10-30 · First seen 2026-11-12

General AI

Abstract

Human working memory lets people maintain and update information. Large language models (LLMs) face a related demand when tracking updates within a single prompt without writing out intermediate states, yet existing tests differ in their constraints on written steps. We introduce a working-memory-inspired probe of cumulative state tracking with program-computed answers, 3 to 100 updates, and final-state, intermediate-state, or multi-target queries. It compares protocols on identical items. Across 18 LLMs, accuracy is 84.9% on single-step arithmetic but 31.7% on core items requiring one number after three to seven updates. From 50 updates on, no model exceeds 10% accuracy. Core accuracy reaches 88.0% under a larger budget and instructions to write steps and place the answer on the last line, mainly due to the instruction changes rather than the budget. Core accuracy also correlates positively with public Berkeley Function Calling Leaderboard scores. These findings show that strong single-step arithmetic does not ensure reliable cumulative tracking and that output protocols substantially change measured accuracy. We release the generator, items, scoring code, and responses for reproducible comparisons across depths, query types, and protocols.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
soon
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@misc{hou2026wmip,
  title = {WMIP: A Working-Memory-Inspired Programmable Probe of Cumulative State Tracking in Large Language Models},
  author = {Dengzhe Hou and Lingyu Jiang and Deng Li and Zirui Li and Fangzhou Lin and Kazunori D Yamada},
  year = {2026},
  abstract = {Human working memory lets people maintain and update information. Large language models (LLMs) face a related demand when tracking updates within a single prompt without writing out intermediate states, yet existing tests differ in their constraints on written steps. We introduce a working-memory-inspired probe of cumulative state tracking with program-computed answers, 3 to 100 updates, and final-state, intermediate-state, or multi-target queries. It compares protocols on identical items. Acros},
  url = {https://huggingface.co/papers/2603.27343},
  keywords = {working memory, cumulative state tracking, large language models, multi-target queries, Berkeley Function Calling Leaderboard, code available, huggingface daily},
  eprint = {2603.27343},
  archiveprefix = {arXiv},
}

Metadata

{}