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

arov

-0.75

 SERVICE

-0.68

CHO

-0.64

arak

-0.62

Sport

-0.62

)]

-0.61

èĢħ

-0.60

 CONTROL

-0.60

adj

-0.60

 CONTR

-0.59

POSITIVE LOGITS

odore

0.80

Ware

0.75

isoft

0.74

Doe

0.70

ldon

0.67

Nah

0.62

bes

0.62

 desper

0.60

olar

0.60

mire

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.