INDEX

Explanations

titles or names of media works

oai_token-act-pair · gpt-3.5-turbo

media-related terms, specifically referring to films, songs, and other entertainment content

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

entimes

-0.66

MEN

-0.63

iland

-0.62

 abound

-0.58

bos

-0.56

bred

-0.56

ruciating

-0.56

cience

-0.54

yrics

-0.54

POSITIVE LOGITS

imester

0.81

akedown

0.70

alion

0.66

 EVER

0.63

 hurdle

0.63

iaz

0.62

 imaginable

0.62

aneers

0.61

 volley

0.60

 arrives

0.60

Activations Density 0.239%

titles or names of media works

media-related terms, specifically referring to films, songs, and other entertainment content

No Comments

No Known Activations