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

bps

-0.67

akis

-0.67

oshenko

-0.66

icians

-0.65

SQL

-0.65

 artisan

-0.64

 Nigerian

-0.63

Wh

-0.63

 Palestinians

-0.62

ician

-0.62

POSITIVE LOGITS

INC

0.66

CVE

0.62

Reviewer

0.61

MRI

0.61

DEC

0.61

isol

0.61

spons

0.60

âĿ

0.60

 Nightmares

0.59

wake

0.59

Activations Density 0.000%

No Known Activations

This feature has no known activations.