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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 CLIENT

-0.77

-0.69

asma

-0.68

lisher

-0.67

interstitial

-0.66

aligned

-0.66

rice

-0.64

DragonMagazine

-0.62

=#

-0.62

 Influ

-0.62

POSITIVE LOGITS

 heating

0.71

rament

0.65

 Fathers

0.65

ologne

0.63

 refuel

0.61

handle

0.60

 probable

0.59

 dope

0.58

 unbeaten

0.58

Ore

0.57

Activations Density 0.000%

No Known Activations

This feature has no known activations.