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

tracking

-0.74

 Adin

-0.71

hander

-0.67

_-

-0.65

atten

-0.64

Nusra

-0.64

omo

-0.62

atin

-0.62

ILY

-0.62

OM

-0.60

POSITIVE LOGITS

ciplinary

0.80

 Books

0.75

alla

0.65

Cop

0.64

 welf

0.64

ventions

0.63

 Boot

0.62

naissance

0.60

uala

0.60

 Horizons

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.