INDEX

Explanations

words related to movie titles or film-related terms

oai_token-act-pair · gpt-3.5-turbo

mentions of specific geographic locations or countries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

tnc

-0.72

edIn

-0.70

 helm

-0.68

blu

-0.63

 Strat

-0.63

 Reson

-0.63

 Thur

-0.62

 pretext

-0.61

 opio

-0.61

 Heads

-0.60

POSITIVE LOGITS

pad

0.84

abad

0.83

mans

0.82

tub

0.79

afia

0.79

rost

0.76

edition

0.75

icz

0.75

pour

0.73

trump

0.71

Activations Density 0.000%

words related to movie titles or film-related terms

mentions of specific geographic locations or countries

No Comments

No Known Activations