INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

aucas

-0.73

 beneficiary

-0.71

animous

-0.69

Zur

-0.69

 Malik

-0.68

VIDIA

-0.68

milo

-0.68

efully

-0.66

 graz

-0.66

erest

-0.65

POSITIVE LOGITS

 Mansion

0.78

é¾įå

0.77

keeping

0.74

nant

0.68

 Quit

0.68

skirts

0.67

adding

0.66

 Surveillance

0.62

VPN

0.62

ents

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.