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

haar

-0.66

hner

-0.65

allah

-0.61

OTO

-0.59

amiya

-0.59

ichever

-0.59

gew

-0.58

 stret

-0.58

00200000

-0.58

ersen

-0.58

POSITIVE LOGITS

Cub

0.71

amen

0.66

IQ

0.65

Connector

0.65

];

0.60

gram

0.60

Rod

0.60

eness

0.60

Material

0.60

atron

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.