INDEX

Explanations

names of political figures, such as "Cheney" and "Rove"

oai_token-act-pair · gpt-3.5-turbo

references to political figures, specifically neoconservatives and Cheney

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

©¶æ¥µ

-0.84

ezvous

-0.77

unal

-0.75

«ĺ

-0.73

phabet

-0.72

 Seller

-0.72

 Pixie

-0.71

alach

-0.69

Ķ

-0.69

Ship

-0.68

POSITIVE LOGITS

ervative

1.18

enegger

1.07

ervatives

0.89

feld

0.88

 Rove

0.84

 Cheney

0.80

 advisers

0.77

 memos

0.76

oleon

0.75

 Warfare

0.74

Activations Density 0.026%

names of political figures, such as "Cheney" and "Rove"

references to political figures, specifically neoconservatives and Cheney

No Comments

No Known Activations