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

 pulse

-0.66

wash

-0.64

 Vance

-0.63

 Colin

-0.60

olk

-0.59

uke

-0.59

perture

-0.58

rup

-0.57

tip

-0.57

espie

-0.57

POSITIVE LOGITS

 "$:/

0.87

 Flavoring

0.80

Sov

0.79

 divid

0.78

¶ħ

0.77

è¦ļéĨĴ

0.75

ģĸ

0.73

ographers

0.70

ç«

0.69

[&

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.