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

 Dynasty

-0.69

NRL

-0.67

 Idol

-0.67

DX

-0.66

ELD

-0.65

 Probe

-0.64

 Monitor

-0.63

 Bearing

-0.63

 Pistol

-0.62

 newsletters

-0.62

POSITIVE LOGITS

ndra

1.04

zzi

0.80

anqu

0.79

okers

0.79

ilib

0.78

oker

0.77

ageddon

0.77

iqueness

0.77

aturally

0.76

ammy

0.76

Activations Density 0.000%

No Known Activations

This feature has no known activations.