INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Piper

-0.75

 Sinai

-0.67

 Cooke

-0.64

 Instruments

-0.62

Alv

-0.59

chi

-0.59

 Fountain

-0.59

 Calder

-0.59

 Odyssey

-0.58

 Jobs

-0.58

POSITIVE LOGITS

Fed

0.77

 impe

0.76

Ĭ±

0.75

erb

0.74

irlf

0.71

 VIDEOS

0.70

anmar

0.68

motion

0.67

edia

0.66

edition

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.