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

 condu

-0.68

SDL

-0.63

 cess

-0.60

coc

-0.60

 Valve

-0.60

tarian

-0.59

itialized

-0.59

 capita

-0.59

 treaties

-0.58

 interpol

-0.58

POSITIVE LOGITS

fall

0.74

bold

0.74

birds

0.73

Bird

0.72

anu

0.72

bats

0.71

BUG

0.71

ouf

0.71

WARD

0.70

Ward

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.