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

 Janeiro

-0.80

 Eden

-0.78

ichick

-0.78

 Lima

-0.72

 Dare

-0.71

anwhile

-0.71

hower

-0.68

celona

-0.66

nesday

-0.66

Os

-0.64

POSITIVE LOGITS

Interest

0.82

à¨

0.78

Known

0.77

========

0.74

Disc

0.71

Experts

0.71

LIB

0.71

FactoryReloaded

0.71

 Factors

0.70

Alternatively

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.