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

riott

-0.73

soType

-0.71

SPONSORED

-0.67

 disguise

-0.64

fourth

-0.63

maps

-0.62

 suspic

-0.62

 separ

-0.61

mercial

-0.61

DATA

-0.61

POSITIVE LOGITS

âĵĺ

0.79

affe

0.72

ashington

0.71

hra

0.68

 Hamm

0.68

ert

0.66

ober

0.65

iazep

0.65

hesda

0.64

ber

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.