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

pace

-0.73

 Giul

-0.72

paces

-0.70

 Camer

-0.66

RUN

-0.63

 Sprint

-0.62

 Diesel

-0.61

 Anthem

-0.60

heed

-0.60

 Wally

-0.59

POSITIVE LOGITS

olkien

0.84

ivated

0.74

ardless

0.73

imated

0.72

wikipedia

0.71

ablishment

0.71

bread

0.67

equ

0.66

Ã§a

0.65

isted

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.