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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ischer

-0.73

NBN

-0.67

âĸ¬

-0.67

 Mahjong

-0.66

Reloaded

-0.64

 Simulator

-0.64

Kap

-0.64

Rah

-0.64

 Cheong

-0.62

Sys

-0.62

POSITIVE LOGITS

agu

0.80

emale

0.78

ounty

0.69

toggle

0.68

agues

0.67

earance

0.65

ocious

0.64

uckle

0.64

 Gorge

0.63

adies

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.