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

 Scotia

-0.80

########

-0.67

 Species

-0.64

IAS

-0.64

 adapted

-0.62

iry

-0.62

 Goose

-0.62

ients

-0.62

 gastro

-0.61

neum

-0.60

POSITIVE LOGITS

kered

0.73

clinton

0.71

htaking

0.70

maxwell

0.67

inventoryQuantity

0.65

politics

0.65

 politics

0.64

electric

0.63

picking

0.63

hillary

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.