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

é¾įå

-0.81

 entitle

-0.71

hiba

-0.68

 enrich

-0.67

HQ

-0.63

peak

-0.62

stim

-0.62

oku

-0.62

 nostalgic

-0.61

wana

-0.61

POSITIVE LOGITS

met

0.75

locks

0.72

"]=>

0.71

ebus

0.66

asser

0.64

Ore

0.64

rained

0.63

 Deity

0.62

Naz

0.62

ENTS

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.