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

 skelet

-0.71

lihood

-0.66

 uncomp

-0.65

 affirmative

-0.64

 unchecked

-0.63

thinkable

-0.60

 citiz

-0.60

ebin

-0.60

 unfavorable

-0.59

ī

-0.59

POSITIVE LOGITS

Lost

0.76

isite

0.75

Davis

0.72

osponsors

0.71

sbm

0.70

Lab

0.70

mat

0.69

irlf

0.66

Detect

0.66

ete

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.