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

ayne

-0.82

uthor

-0.80

auga

-0.80

rights

-0.79

itizen

-0.78

ondo

-0.78

undy

-0.74

oxide

-0.74

eln

-0.74

anchester

-0.74

POSITIVE LOGITS

sub

0.94

 AUTH

0.78

 treat

0.77

 fetch

0.70

 IEEE

0.67

SUP

0.67

 doub

0.66

 trop

0.65

 Muse

0.64

UL

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.