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

undai

-0.72

hus

-0.72

psons

-0.69

apore

-0.69

apons

-0.69

acterial

-0.69

este

-0.67

ãĥ¼ãĤ¯

-0.67

unda

-0.66

pse

-0.65

POSITIVE LOGITS

 Ashe

0.68

 Ways

0.64

BET

0.63

Buff

0.62

ACY

0.62

DERR

0.61

 buds

0.60

âĢ¢âĢ¢

0.60

EStream

0.60

hower

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.