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

 hurd

-0.81

 indo

-0.66

staking

-0.66

 prospect

-0.64

leans

-0.63

 millennium

-0.63

 bour

-0.63

 Mehran

-0.62

verts

-0.61

 programmes

-0.61

POSITIVE LOGITS

ãĤ´ãĥ³

0.89

notations

0.83

ļ

0.77

Ń·

0.77

ICLE

0.71

terness

0.71

":"/

0.68

ģ

0.68

 STATS

0.68

Oo

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.