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

 Ambro

-0.71

Dialogue

-0.71

 vigilance

-0.69

 Olsen

-0.68

otten

-0.67

 opio

-0.64

 Carnage

-0.63

Alert

-0.63

©¶æ

-0.62

 unden

-0.62

POSITIVE LOGITS

pez

0.91

="/

0.89

ppo

0.75

Indian

0.75

="

0.74

gencies

0.71

frame

0.68

eri

0.67

heads

0.67

]=

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.