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

\\\\\\\\

-0.76

 permitting

-0.67

 masculinity

-0.66

 kindness

-0.64

'';

-0.63

 cane

-0.62

 assignments

-0.62

 paddle

-0.62

 LIFE

-0.61

 judgment

-0.61

POSITIVE LOGITS

ree

0.80

debug

0.77

emis

0.72

rie

0.68

rell

0.67

Streamer

0.67

ren

0.66

aspx

0.65

ischer

0.65

leep

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.