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

 Western

-0.76

 bluff

-0.75

 Eastern

-0.71

 Willow

-0.68

 Twin

-0.65

 Oregon

-0.65

 Southwest

-0.65

 paddle

-0.65

 Delaware

-0.64

 Northwest

-0.64

POSITIVE LOGITS

bugs

0.79

atis

0.79

 CONTR

0.78

igers

0.77

onds

0.76

issues

0.75

guards

0.74

awi

0.73

esome

0.71

ributes

0.71

Activations Density 0.000%

No Known Activations

This feature has no known activations.