INDEX

Explanations

phrases involving legal charges and accusations

oai_token-act-pair · gpt-3.5-turbo

references to legal accusations and indictments

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

Negative Logits

birth

-0.78

olen

-0.72

 grass

-0.71

 leisure

-0.68

 Kers

-0.68

perm

-0.66

hall

-0.65

irst

-0.64

hair

-0.63

events

-0.63

POSITIVE LOGITS

 indictment

1.38

 indict

0.89

eering

0.89

 indicted

0.88

 levied

0.77

 affidavit

0.75

mination

0.73

 dismissing

0.73

 accuses

0.72

 alleges

0.71

Activations Density 0.009%

phrases involving legal charges and accusations

references to legal accusations and indictments

No Known Activations