INDEX

Explanations

mentions of specific figures or entities

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "sci-fi" or science fiction

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

 quizz

-0.68

İĭ

-0.68

 clipboard

-0.65

 Hyper

-0.61

NXT

-0.60

lain

-0.59

cont

-0.59

poke

-0.56

isner

-0.56

oid

-0.55

POSITIVE LOGITS

fi

1.30

Fi

1.01

eld

0.86

 ILCS

0.82

endish

0.79

olini

0.76

renheit

0.76

ennes

0.76

agascar

0.76

velength

0.75

Activations Density 0.010%

mentions of specific figures or entities

references to the concept of "sci-fi" or science fiction

No Comments

No Known Activations