INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

estinal

-0.73

risis

-0.72

aepernick

-0.69

oliberal

-0.67

ayne

-0.65

cod

-0.65

uncture

-0.65

agonist

-0.65

iltration

-0.65

inguished

-0.64

POSITIVE LOGITS

Downloadha

0.73

åĤ

0.71

 Sawyer

0.71

 Chambers

0.71

 Cald

0.69

 Hath

0.68

ãĥķãĤ¡

0.67

 âĢº

0.66

éļ

0.66

istics

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.