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

rique

-0.77

ĸļ

-0.75

 reperto

-0.75

Interstitial

-0.73

Methods

-0.72

»Ĵ

-0.71

CPC

-0.71

UGC

-0.70

icons

-0.68

querque

-0.68

POSITIVE LOGITS

hen

0.74

Â°

0.67

hend

0.66

arcer

0.66

 Mort

0.65

Gas

0.63

 Plot

0.62

atherine

0.61

gaard

0.61

rend

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.