AI Navigate
インサイト最新記事AI大全カオスマップ

Across the Levels of Analysis: Explaining Predictive Processing in Humans Requires More Than Machine-Estimated Probabilities

arXiv cs.CL / 4/13/2026

💬 OpinionIdeas & Deep AnalysisModels & Research
Read original →
共有:

Key Points

  • The paper uses Marr’s levels of analysis to critique two common claims linking language processing to probabilistic prediction and suggesting psycholinguistic advances depend on large language models (LLMs).
  • It argues that explaining human predictive processing requires more than machine-estimated probabilities derived from language models.
  • The authors extend the discussion on how prediction is central to language processing, while questioning simplistic mappings between model likelihoods and human cognition.
  • The work proposes future research directions that integrate strengths of LLMs with psycholinguistic models to better account for human language behavior.

Abstract

Under the lens of Marr's levels of analysis, we critique and extend two claims about language models (LMs) and language processing: first, that predicting upcoming linguistic information based on context is central to language processing, and second, that many advances in psycholinguistics would be impossible without large language models (LLMs). We further outline future directions that combine the strengths of LLMs with psycholinguistic models.

関連おすすめサービス

※当サイトはアフィリエイト広告を利用しています
🎧

Notta搭載AI議事録イヤホン ZENCHORD1

AI時代の仕事術。Notta搭載で会議の議事録を自動生成するスマートイヤホン。

🎙️

AI搭載ボイスレコーダー Plaud

世界100万人が愛用。AIで文字起こし・要約を自動化するボイスレコーダー。

🖼️

画像高画質化AIツール Aiarty Image Enhancer

AIで画像を高画質化。写真・イラストを簡単にアップスケール。