Explaining Large Language Models Decisions Using Shapley Values · HackerPapers discussions

Explaining Large Language Models Decisions Using Shapley Values — mentioned by behnamoh

Fri, 20 Jun 2025 14:33:00 GMT

Yes, this paper says exactly what you talked about: https://arxiv.org/abs/2404.01332

Paper and discussion history. A mention is not an endorsement.

Explaining Large Language Models Decisions Using Shapley Values — HN submission

Sat, 28 Dec 2024 00:44:18 GMT

Explaining Large Language Models Decisions Using Shapley Values

89 points; 19 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Explaining Large Language Models Decisions Using Shapley Values — HN submission

Sun, 10 Nov 2024 16:38:11 GMT

LLM outputs explained using Game Theory

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Explaining Large Language Models Decisions Using Shapley Values — HN submission

Fri, 02 Aug 2024 04:37:08 GMT

Interpreting LLM outputs reveals the "token noise" effect

2 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Explaining Large Language Models Decisions Using Shapley Values — mentioned by behnamoh

Sun, 19 May 2024 00:47:23 GMT

Whenever "AI influencers" say "prompt engineering is dead", I realize they are full of bs. If anything, prompt engineering has become more nuanced, down to the choice of each token: https://arxiv.org/abs/2404.01332

Paper and discussion history. A mention is not an endorsement.

Explaining Large Language Models Decisions Using Shapley Values — HN submission

Wed, 03 Apr 2024 21:06:26 GMT

LLMs and the "token noise" effect

1 point; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.

Explaining Large Language Models Decisions Using Shapley Values — HN submission

Wed, 03 Apr 2024 02:16:14 GMT

Wait, It's All Token Noise? Always Has Been: Interpreting LLM Behavior

2 points; 0 discussion replies on HN. Paper and source history. A submission is not an endorsement.