A Formal Limitation on Learning Human Language From Textual Corpora
Ranking
Overall
78
Content
95
Popularity
37
Observed public metrics from 1 member.
Merged summary
TL;DR - This paper derives information-theoretic limits on recovering a speaker’s intended meaning from text alone, including from LLM hidden states. It shows that ambiguity requiring extralinguistic context cannot be eliminated by better representations, more training data, or additional supervision.
- Models language as a joint distribution over meanings, contexts, and utterances, then bounds a decoder’s probability of recovering intended meaning.
- Separates uncertainty into an irreducible component and a component resolvable only through extralinguistic context.
- Applies to any text featurizer and to both discrete and continuous meaning spaces.
- Experiments on artificial languages, Mandarin zero-pronoun resolution, and color reference provide empirical support for the theory.
Sources (1)
A Formal Limitation on Learning Human Language From Textual Corpora
Public signals
Semantic Scholar citations 0 · Semantic Scholar influential citations 0
TL;DR - This paper derives information-theoretic limits on recovering a speaker’s intended meaning from text alone, including from LLM hidden states. It shows that ambiguity requiring extralinguistic context cannot be eliminated by better representations, more training data, or additional supervision.
- Models language as a joint distribution over meanings, contexts, and utterances, then bounds a decoder’s probability of recovering intended meaning.
- Separates uncertainty into an irreducible component and a component resolvable only through extralinguistic context.
- Applies to any text featurizer and to both discrete and continuous meaning spaces.
- Experiments on artificial languages, Mandarin zero-pronoun resolution, and color reference provide empirical support for the theory.