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

#$

-0.72

 Inher

-0.62

ģĸ

-0.62

 Disability

-0.62

 insur

-0.62

obs

-0.61

ype

-0.59

 Twain

-0.58

 coping

-0.58

RL

-0.57

POSITIVE LOGITS

theless

0.80

olson

0.75

cham

0.73

pher

0.71

tein

0.70

zbollah

0.69

ascus

0.69

mony

0.68

oise

0.68

HER

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.