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

aire

-0.72

 poles

-0.67

agy

-0.65

WAY

-0.63

 fortun

-0.61

 vigil

-0.61

 wisely

-0.61

 pole

-0.61

ship

-0.57

 Fein

-0.57

POSITIVE LOGITS

Downloadha

0.84

imony

0.77

 âĢİ

0.74

SPONSORED

0.70

 ..."

0.68

emale

0.65

duction

0.64

artment

0.64

gradient

0.64

Scient

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.