INDEX

Explanations

the word "Reid" with varying levels of activation

oai_token-act-pair · gpt-3.5-turbo

mentions of the individual named Reid

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

andise

-0.99

omaly

-0.81

ruciating

-0.78

teenth

-0.70

tesy

-0.68

oppable

-0.68

 dehyd

-0.66

ugal

-0.65

eatures

-0.65

 exha

-0.63

POSITIVE LOGITS

 Reid

0.85

burgh

0.84

sburg

0.84

igan

0.77

ety

0.73

 Mysteries

0.72

les

0.71

bro

0.71

lings

0.70

 Feinstein

0.70

Activations Density 0.010%

the word "Reid" with varying levels of activation

mentions of the individual named Reid

No Comments

No Known Activations