INDEX

Explanations

phrases related to political controversy and conflict

oai_token-act-pair · gpt-3.5-turbo

references to political conflict and personal attacks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

BuyableInstoreAndOnline

-0.78

emonium

-0.73

 Nanto

-0.66

iHUD

-0.65

ARB

-0.62

ITNESS

-0.61

 juggling

-0.61

 shuffle

-0.61

 Boom

-0.60

uned

-0.59

POSITIVE LOGITS

 indiscrim

0.96

 unfairly

0.93

 unnecessarily

0.93

 lest

0.89

 because

0.89

 unjust

0.87

 outright

0.85

selves

0.83

by

0.83

 innoc

0.83

Activations Density 0.370%

phrases related to political controversy and conflict

references to political conflict and personal attacks

No Comments

No Known Activations