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

rend

-0.73

bery

-0.72

anan

-0.70

review

-0.69

 Rober

-0.69

andowski

-0.69

arist

-0.68

izes

-0.68

nets

-0.67

endor

-0.67

POSITIVE LOGITS

 pulses

0.78

 filament

0.71

 chrom

0.69

LH

0.66

CBD

0.66

 phased

0.65

 heartbeat

0.64

 periodic

0.64

 Plasma

0.63

 cured

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.