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

Jud

-0.68

ept

-0.67

 madness

-0.62

lier

-0.61

 prest

-0.58

 Rampage

-0.57

zon

-0.56

 strugg

-0.56

 Alzheimer

-0.56

 Magnetic

-0.56

POSITIVE LOGITS

icum

0.79

ourced

0.77

ICE

0.74

SN

0.73

netflix

0.73

aiman

0.72

pmwiki

0.69

arov

0.68

license

0.68

orst

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.