INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

iets

-0.89

hess

-0.89

ogue

-0.86

okane

-0.77

usp

-0.74

bye

-0.73

apsed

-0.71

clinton

-0.70

erity

-0.70

minster

-0.69

POSITIVE LOGITS

Ń·

0.76

 mosquit

0.71

 generic

0.70

ãĥ³ãĤ¸

0.69

 selecting

0.68

 Commando

0.66

 fetch

0.66

 Pricing

0.64

nab

0.63

 Protocol

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.