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

ãĥĨãĤ£

-0.84

stocks

-0.67

close

-0.65

ezvous

-0.65

coe

-0.64

ACTED

-0.64

utm

-0.63

itive

-0.62

amia

-0.62

Chip

-0.61

POSITIVE LOGITS

 Norn

0.86

 Primordial

0.66

 Eden

0.65

 Teaching

0.64

ç¥ŀ

0.64

 Horses

0.63

 Anxiety

0.63

 Sere

0.63

 Princ

0.62

 Nurs

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.