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

 Canaver

-0.80

 Sands

-0.79

Eag

-0.75

 McMaster

-0.74

 Byrne

-0.73

 Foley

-0.73

 Gillespie

-0.71

 Wynne

-0.70

 Spicer

-0.70

 Winc

-0.70

POSITIVE LOGITS

 intensive

0.69

ulia

0.69

restrial

0.67

pil

0.66

oven

0.65

veh

0.65

 unreal

0.63

xual

0.62

igmat

0.62

cient

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.