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

»Ĵ

-0.92

bernatorial

-0.77

vironments

-0.77

merce

-0.77

 Avalanche

-0.70

alach

-0.70

å§«

-0.70

bably

-0.70

 trave

-0.69

>>\

-0.69

POSITIVE LOGITS

catentry

0.71

ocratic

0.69

 prize

0.69

oning

0.65

seq

0.64

aire

0.64

oo

0.63

wr

0.61

gency

0.60

 inspections

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.