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

insula

-0.68

 Either

-0.68

mast

-0.66

:(

-0.66

oca

-0.65

 Participation

-0.64

cup

-0.63

osterone

-0.62

ilation

-0.62

ulkan

-0.60

POSITIVE LOGITS

 secondly

0.78

 consequently

0.72

romeda

0.71

kus

0.71

alus

0.70

 devast

0.66

thur

0.63

 accelerate

0.63

 conductor

0.62

hester

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.