INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ohyd

-0.70

inth

-0.69

bench

-0.68

ixel

-0.67

ipp

-0.64

 Dome

-0.63

 Cheong

-0.62

aver

-0.62

ozo

-0.62

prus

-0.61

POSITIVE LOGITS

¥µ

0.82

Interstitial

0.81

EngineDebug

0.77

 behavi

0.69

IRD

0.68

ç·

0.67

Sov

0.66

Reviewer

0.65

Software

0.63

 transl

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.