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

halla

-0.92

chio

-0.85

rities

-0.83

etheless

-0.81

atorial

-0.78

umerable

-0.76

kson

-0.76

isine

-0.75

erity

-0.74

eteria

-0.73

POSITIVE LOGITS

0.75

IE

0.66

 guaranteed

0.65

 rooted

0.64

 electr

0.62

 seaf

0.61

 unstable

0.60

yg

0.59

rain

0.58

ag

0.58

Activations Density 0.000%

No Known Activations

This feature has no known activations.