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

 psychiat

-0.68

bring

-0.68

oz

-0.66

chel

-0.65

borgh

-0.65

visors

-0.64

oka

-0.64

ulet

-0.63

oglobin

-0.62

 territ

-0.62

POSITIVE LOGITS

 rigging

0.71

 Egyptian

0.67

 Madagascar

0.67

abouts

0.66

 Egyptians

0.65

 ruins

0.65

 MIDI

0.65

onal

0.64

maxwell

0.63

 Egypt

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.