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

ierrez

-0.73

theless

-0.71

×Ļ×

-0.69

rang

-0.69

nel

-0.69

ãĥĦ

-0.68

||||

-0.67

ModLoader

-0.67

manship

-0.66

POSITIVE LOGITS

 Liberties

0.84

tarians

0.84

 systemd

0.77

aries

0.71

keyes

0.68

AoE

0.68

 libertarians

0.67

 EVEN

0.67

erto

0.67

tarian

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.