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

STATE

-0.69

icum

-0.64

STR

-0.63

US

-0.62

strate

-0.62

ected

-0.62

FR

-0.61

BY

-0.61

ounters

-0.60

 FILE

-0.60

POSITIVE LOGITS

fell

1.13

 veter

0.73

 nodd

0.71

 juven

0.70

 destro

0.70

 Jagu

0.69

 tiss

0.68

gaard

0.65

 elders

0.64

 Moroc

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.