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

ĸļ

-0.78

ctuary

-0.75

arie

-0.74

icter

-0.70

Balt

-0.69

 treaties

-0.69

rals

-0.68

arthed

-0.67

gettable

-0.66

 warr

-0.66

POSITIVE LOGITS

opausal

0.79

YA

0.69

 propagate

0.61

olves

0.61

ulf

0.61

ç¥ŀ

0.60

 characterize

0.60

hat

0.60

UTC

0.60

hai

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.