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

olver

-0.80

ominated

-0.79

 abst

-0.73

ighed

-0.73

vr

-0.72

 exorc

-0.72

inker

-0.71

ingen

-0.70

rawdownloadcloneembedreportprint

-0.68

 reactionary

-0.67

POSITIVE LOGITS

 Macy

0.78

 Chaff

0.78

 Designs

0.77

éĹ

0.75

æ©Ł

0.74

Taj

0.72

İĭ

0.71

 Fields

0.69

 Frie

0.67

 Hilton

0.67

Activations Density 0.000%

No Known Activations

This feature has no known activations.