INDEX

Explanations

phrases indicating spatial or relational distance

oai_token-act-pair · gpt-3.5-turbo

references to spatial relationships or connections

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ERY

-0.73

bye

-0.73

Shape

-0.69

acea

-0.69

nit

-0.69

itcher

-0.66

notation

-0.66

aston

-0.66

iam

-0.65

claimer

-0.64

POSITIVE LOGITS

 sexes

1.73

two

1.51

 extremes

1.30

 genders

1.27

 halves

1.15

 parties

1.09

 ages

1.08

 dots

1.06

 three

1.01

 worlds

1.01

Activations Density 0.165%

phrases indicating spatial or relational distance

references to spatial relationships or connections

No Comments

No Known Activations