INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

orno

-0.73

oldown

-0.72

 lectures

-0.70

audi

-0.70

hiba

-0.69

opian

-0.69

intage

-0.68

 enjoyment

-0.68

 millenn

-0.68

 Editors

-0.67

POSITIVE LOGITS

 ÃĹ

0.77

â

0.73

ISA

0.71

ãģ®éŃĶ

0.69

ARP

0.68

KO

0.66

PAN

0.65

MORE

0.64

Aim

0.64

PL

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.