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

 adolesc

-0.72

 TAMADRA

-0.70

 carc

-0.68

 Flame

-0.65

[|

-0.65

 Mechdragon

-0.64

 BALL

-0.64

 Cache

-0.64

Oro

-0.63

inflamm

-0.63

POSITIVE LOGITS

hend

0.74

Pr

0.71

hs

0.70

politics

0.69

adian

0.68

Shares

0.68

TW

0.68

PIN

0.67

Pict

0.67

Secure

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.