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

ure

-0.71

 grounding

-0.69

 dÃ©

-0.64

ining

-0.64

aire

-0.63

ation

-0.62

tto

-0.62

ty

-0.60

stress

-0.59

 reacting

-0.59

POSITIVE LOGITS

agher

1.02

alion

0.98

 Hemp

0.87

acl

0.81

halla

0.80

nels

0.79

ModLoader

0.75

EH

0.72

entanyl

0.71

jri

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.