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

 surn

-0.69

 dearly

-0.68

 Peck

-0.64

 Pirate

-0.64

pie

-0.61

 bonuses

-0.60

Khe

-0.60

 maths

-0.59

 Yellow

-0.58

 gigs

-0.58

POSITIVE LOGITS

idity

0.82

framework

0.82

ohm

0.81

ansion

0.77

iste

0.76

0.74

aro

0.74

apon

0.73

isco

0.73

à¼

0.73

Activations Density 0.000%

No Known Activations

This feature has no known activations.