INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

cius

-0.70

azes

-0.70

 fram

-0.68

yss

-0.67

ington

-0.67

sab

-0.67

é»Ĵ

-0.67

ares

-0.66

isan

-0.66

bucks

-0.66

POSITIVE LOGITS

ailability

0.74

])

0.67

added

0.65

arten

0.62

 covered

0.61

Ka

0.61

chnology

0.60

 headline

0.59

awed

0.59

 snug

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.