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

]-

-0.70

 Conquest

-0.66

 Mellon

-0.65

 Scroll

-0.64

 Ital

-0.63

Sop

-0.63

 Hispan

-0.63

 Aval

-0.63

 Samoa

-0.62

 scrimmage

-0.61

POSITIVE LOGITS

nex

0.73

backer

0.71

nesota

0.66

vib

0.65

uden

0.65

erk

0.65

stro

0.63

ocial

0.63

uto

0.63

san

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.