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

rina

-0.76

opped

-0.70

 TAMADRA

-0.69

umsy

-0.66

inic

-0.66

iatrics

-0.62

ryce

-0.62

cluded

-0.61

aughtered

-0.61

tered

-0.61

POSITIVE LOGITS

cean

0.70

 Bing

0.68

idth

0.67

clone

0.67

VERTISEMENT

0.66

vernment

0.65

GMT

0.63

DragonMagazine

0.62

tti

0.62

TPS

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.