INDEX

Explanations

mentions of individuals being openly part of a specific group or having a specific characteristic

oai_token-act-pair · gpt-3.5-turbo

instances of the word "openly" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Summoner

-0.76

akeru

-0.75

nesota

-0.72

ĸļ

-0.70

ocre

-0.68

 Tune

-0.66

 Enchant

-0.64

ammy

-0.64

 Karma

-0.64

anan

-0.62

POSITIVE LOGITS

obin

0.82

odox

0.80

 openly

0.74

lishing

0.74

 disreg

0.74

lit

0.73

 confess

0.72

 defied

0.72

 expressed

0.71

 profess

0.71

Activations Density 0.011%

mentions of individuals being openly part of a specific group or having a specific characteristic

instances of the word "openly" in various contexts

No Comments

No Known Activations