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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 messenger

-0.72

 camel

-0.65

 Presents

-0.63

 hoops

-0.63

ingen

-0.63

 Hitchcock

-0.61

 airs

-0.60

 accent

-0.60

 Foot

-0.59

 Express

-0.57

POSITIVE LOGITS

GS

1.29

AGES

0.84

Anyway

0.77

Num

0.75

ADRA

0.75

ETS

0.74

GROUP

0.73

STD

0.72

FY

0.71

perty

0.70

Activations Density 0.000%

No Known Activations

This feature has no known activations.