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

MQ

-0.81

 Hail

-0.71

éģ

-0.70

-0.69

Nil

-0.67

oter

-0.66

eering

-0.65

 Riding

-0.64

å¦

-0.64

 River

-0.63

POSITIVE LOGITS

adobe

0.82

chers

0.71

anyl

0.69

displayText

0.68

maxwell

0.66

spr

0.65

ensitive

0.63

anced

0.63

psey

0.62

comed

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.