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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

EStreamFrame

-0.76

abase

-0.75

 Broadcasting

-0.67

 whistleblowers

-0.66

 informants

-0.64

amount

-0.63

isms

-0.62

Esp

-0.62

doms

-0.61

Ring

-0.60

POSITIVE LOGITS

lde

0.68

gra

0.66

 mate

0.66

ERE

0.66

Redditor

0.64

esis

0.62

 associate

0.62

ALSE

0.62

 Scor

0.61

SLI

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.