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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

fen

-0.77

 Bulgar

-0.71

SU

-0.69

emis

-0.67

 Bakr

-0.65

 execut

-0.65

malink

-0.63

 Brach

-0.62

uana

-0.62

ZE

-0.62

POSITIVE LOGITS

dies

0.65

 Islands

0.64

ict

0.64

cipline

0.62

eston

0.62

camera

0.60

ership

0.59

sea

0.58

Ĵ

0.57

 brakes

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.