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

natureconservancy

-0.75

tu

-0.71

 mathemat

-0.70

cffffcc

-0.70

 mamm

-0.67

atorium

-0.64

%%%%

-0.64

ORK

-0.64

wcsstore

-0.63

 Pixie

-0.61

POSITIVE LOGITS

ocratic

0.73

 Electoral

0.70

ificantly

0.67

 Nation

0.66

 Solitaire

0.66

 Green

0.65

ocracy

0.62

 Republic

0.62

 Aging

0.62

 latter

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.