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

 tense

-0.63

Tow

-0.62

 Romeo

-0.62

 outrage

-0.61

 Pound

-0.61

 Tale

-0.60

 theat

-0.59

 Inferno

-0.58

Bis

-0.58

 Bord

-0.57

POSITIVE LOGITS

quickShipAvailable

0.78

nect

0.76

aterasu

0.73

)</

0.73

overe

0.72

orers

0.71

icators

0.70

ojure

0.70

resa

0.69

istance

0.69

Activations Density 0.000%

No Known Activations

This feature has no known activations.