INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

uminati

-0.82

 collapsing

-0.80

 Dying

-0.71

 Invisible

-0.69

 powerless

-0.69

 counting

-0.65

 pyramid

-0.63

 coffin

-0.62

 decoration

-0.61

 dumb

-0.60

POSITIVE LOGITS

hello

0.82

esp

0.74

anta

0.72

bring

0.71

asma

0.70

ãĢĳ

0.67

Steam

0.66

Ì

0.66

arious

0.66

 Strikes

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.