INDEX

Explanations

words related to different types of activities or concepts

oai_token-act-pair · gpt-3.5-turbo

references to various forms of entertainment and their contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

atus

-0.61

xus

-0.59

pload

-0.58

 Ezek

-0.56

milo

-0.55

 nomine

-0.53

DonaldTrump

-0.53

izont

-0.53

-0.51

 seiz

-0.51

POSITIVE LOGITS

 alike

1.29

 respectively

1.01

 industries

0.60

depending

0.58

 depending

0.56

 sectors

0.56

 thereof

0.55

 versa

0.54

 combine

0.54

 modes

0.49

Activations Density 0.464%

words related to different types of activities or concepts

references to various forms of entertainment and their contexts

No Comments

No Known Activations