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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ESE

-0.75

ayers

-0.74

asher

-0.71

ipel

-0.69

/-

-0.69

aved

-0.67

ayer

-0.64

rower

-0.63

days

-0.63

acket

-0.62

POSITIVE LOGITS

netflix

0.84

near

0.81

 near

0.78

ãĤ¤ãĥĪ

0.74

 Loch

0.68

 Karin

0.67

 Niet

0.66

 Knot

0.65

 Kats

0.65

Clar

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.