INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ogl

-0.70

£ı

-0.70

fin

-0.66

Ã±

-0.63

ĻĤ

-0.62

 termin

-0.60

astics

-0.59

 negoti

-0.59

 recomp

-0.59

 confir

-0.57

POSITIVE LOGITS

ampunk

0.71

izabeth

0.70

umblr

0.69

thirst

0.69

law

0.66

 Perkins

0.65

worm

0.65

heimer

0.64

apult

0.64

Article

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.