INDEX

Explanations

references to the media company "Fox"

oai_token-act-pair · gpt-3.5-turbo

mentions of the television network Fox News

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

NetMessage

-0.87

 proble

-0.75

accompan

-0.74

 containers

-0.71

Ñģ

-0.70

sylv

-0.70

PART

-0.69

 transactions

-0.68

à©

-0.68

thood

-0.68

POSITIVE LOGITS

Fox

3.78

Fox

3.14

FOX

2.47

fox

2.23

FOX

2.03

fox

2.02

ABC

1.55

CBS

1.52

NBC

1.48

 Murdoch

1.35

Activations Density 0.030%

references to the media company "Fox"

mentions of the television network Fox News

No Comments

No Known Activations