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

stroke

-0.69

culosis

-0.66

ts

-0.65

internet

-0.64

 stroke

-0.64

 Genetics

-0.62

 Tenth

-0.62

angs

-0.61

astern

-0.61

abyte

-0.61

POSITIVE LOGITS

uper

0.83

IBLE

0.71

ãĤ©

0.69

ero

0.66

Route

0.64

uman

0.64

owered

0.63

tarians

0.62

 vigil

0.61

imated

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.