INDEX

Explanations

words related to controversial topics

oai_token-act-pair · gpt-3.5-turbo

instances of the word "controversial."

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

igi

-0.79

vation

-0.79

vest

-0.78

ynthesis

-0.78

amina

-0.76

á

-0.72

ept

-0.72

ften

-0.72

conom

-0.71

anza

-0.69

POSITIVE LOGITS

 controversial

1.04

 unpopular

0.80

 topics

0.79

 controvers

0.79

 controversies

0.78

 taboo

0.78

 fringe

0.75

roversial

0.74

 contentious

0.71

 fallout

0.71

Activations Density 0.014%

words related to controversial topics

instances of the word "controversial."

No Comments

No Known Activations