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

fred

-0.75

pmwiki

-0.74

 suspic

-0.71

hops

-0.70

zig

-0.70

 WARN

-0.68

 encount

-0.66

querque

-0.65

jar

-0.65

aqu

-0.64

POSITIVE LOGITS

({

0.72

eg

0.70

isine

0.67

([

0.65

 breakdown

0.64

 Dialogue

0.63

going

0.63

 Downs

0.61

iday

0.61

ounds

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.