Representation Engineering: A Top-Down Approach to AI Transparency · HackerPapers discussions

Representation Engineering: A Top-Down Approach to AI Transparency — mentioned by 13point5

Thu, 11 Sep 2025 05:31:28 GMT

I've seen a lot of purple gradient websites made by LLMs and I was curious if we can change a model's favorite color by messing with the activations instead of prompts. So I used Representation Engineering with just 1 pair of contrastive prompts to make Mistral-7B prefer orange as its favorite color. I used the repeng library by Theia to test this out. Next I'm gonna implement it from scratch to understand why this even works. I wonder if we can introduce "taste" into a model with methods like this. The paper is "REPRESENTATION ENGINEERING: A TOP-DOWN APPROACH TO AI TRANSPARENCY". Link…

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

Representation Engineering: A Top-Down Approach to AI Transparency — mentioned by regularfry

Sun, 25 Feb 2024 13:39:43 GMT

Have a read of https://arxiv.org/abs/2310.01405. It describes how an emotional state can be identified as an emergent property of an LLM's activations, and how manipulating that emotional state can affect compliance to requests.

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

Representation Engineering: A Top-Down Approach to AI Transparency — HN submission

Wed, 04 Oct 2023 15:38:43 GMT

Representation Engineering: A Top-Down Approach to AI Transparency

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