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

pedia

-0.75

ipeg

-0.72

etsk

-0.72

ewater

-0.72

checking

-0.70

ðĿ

-0.68

 electors

-0.67

illary

-0.65

DEM

-0.64

gas

-0.64

POSITIVE LOGITS

 Sheikh

0.66

 Noon

0.61

 Chairman

0.59

 tailor

0.59

Mum

0.58

 misinterpret

0.57

Bed

0.57

orns

0.56

 Clock

0.56

 Racer

0.55

Activations Density 0.000%

No Known Activations

This feature has no known activations.