INDEX

Explanations

proper names and entities likely related to a specific location or organization

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names and locations

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

 Spectre

-0.69

Lev

-0.67

Bound

-0.65

 Ruth

-0.64

 Corona

-0.63

Score

-0.63

Ber

-0.62

 Phelps

-0.61

 Berger

-0.61

 Brave

-0.61

POSITIVE LOGITS

sembly

0.92

colm

0.91

ional

0.90

nam

0.89

nas

0.88

ignty

0.87

ctory

0.86

enei

0.86

nis

0.84

iment

0.84

Activations Density 0.022%

proper names and entities likely related to a specific location or organization

proper nouns, particularly names and locations

No Comments

No Known Activations