INDEX

Explanations

statements by authors within research or academic context

oai_token-act-pair · gpt-3.5-turbo

mentions of "authors" in relation to research or scholarly work

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Territories

-0.80

 Corps

-0.73

Gun

-0.71

 Breaker

-0.69

 Oval

-0.68

 Electrical

-0.66

 Labor

-0.65

 Terr

-0.64

SEE

-0.63

Sau

-0.63

POSITIVE LOGITS

hip

1.30

itatively

1.18

paces

1.08

 authors

1.04

uggest

0.97

emis

0.97

essee

0.94

authors

0.93

manship

0.92

pace

0.91

Activations Density 0.014%

statements by authors within research or academic context

mentions of "authors" in relation to research or scholarly work

No Comments

No Known Activations