INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ommel

-0.75

 Sard

-0.69

Squ

-0.69

 Balk

-0.68

ISM

-0.67

ostics

-0.62

 Flavoring

-0.61

 nerv

-0.61

iments

-0.59

ysis

-0.59

POSITIVE LOGITS

iary

0.72

 Chocobo

0.72

 Cummings

0.70

 Volt

0.67

 Omni

0.67

ĳå£«

0.66

ARC

0.65

 Vald

0.65

 Cinderella

0.64

 Shinra

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.