INDEX

Explanations

the word "kin" with varying activations

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

enance

-0.86

 Seym

-0.75

 cryst

-0.72

eas

-0.72

ffic

-0.71

OND

-0.67

CAST

-0.67

inance

-0.65

Ö¼

-0.65

reon

-0.64

POSITIVE LOGITS

etic

1.26

etics

1.25

etically

1.20

wagen

1.10

ned

1.08

ny

1.06

ning

1.06

ners

1.00

esthetic

0.98

creen

0.97

Activations Density 0.027%

the word "kin" with varying activations

No Comments

No Known Activations