INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 substitution

-0.66

 "$:/

-0.60

 ninth

-0.60

jri

-0.59

 monarch

-0.58

pedia

-0.57

 recruit

-0.57

 usur

-0.57

 scholarship

-0.56

 Thrones

-0.56

POSITIVE LOGITS

aud

1.26

urate

0.87

chio

0.81

ums

0.75

cone

0.73

afort

0.70

orum

0.70

auer

0.70

vr

0.70

kson

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.