INDEX

Explanations

phrases related to political figures and events

oai_token-act-pair · gpt-3.5-turbo

references to political figures or discussions about politics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 decomp

-0.72

hift

-0.68

 comr

-0.66

 JPEG

-0.64

cro

-0.61

 scrim

-0.61

ĨĴ

-0.61

 unaccount

-0.58

 Mess

-0.57

 Rivals

-0.56

POSITIVE LOGITS

1.22

ship

1.06

alty

0.95

ufact

0.93

tarian

0.89

sf

0.88

tal

0.88

gins

0.87

ity

0.87

sg

0.86

Activations Density 0.115%

phrases related to political figures and events

references to political figures or discussions about politics

No Comments

No Known Activations