INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

chev

-0.81

rament

-0.72

ventus

-0.72

miss

-0.71

okÃ©

-0.71

plan

-0.69

rome

-0.69

pic

-0.69

 Cheong

-0.68

ano

-0.64

POSITIVE LOGITS

@@

0.81

 partName

0.79

aults

0.73

ĸļ

0.70

ecd

0.70

0.65

 proceeds

0.65

adesh

0.64

guards

0.64

wit

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.