INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

dule

-0.79

estyle

-0.72

othal

-0.72

 Sioux

-0.69

ptin

-0.69

erness

-0.68

aroo

-0.67

ohm

-0.65

ĸļ

-0.65

Mew

-0.65

POSITIVE LOGITS

 confir

0.87

 targ

0.82

 tradem

0.76

 acquaintance

0.70

ESCO

0.68

 destro

0.68

 pled

0.68

 toget

0.67

 trespass

0.66

 arte

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.