INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ule

-0.76

inately

-0.76

orest

-0.70

odiac

-0.68

orney

-0.66

fortune

-0.66

unders

-0.65

veland

-0.65

inent

-0.65

ernand

-0.64

POSITIVE LOGITS

 admitting

0.72

 McGr

0.69

 McDonnell

0.64

vier

0.63

Ts

0.61

BuyableInstoreAndOnline

0.61

 Olson

0.61

 therefore

0.61

Thor

0.60

Ultimately

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.