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

 Simulator

-0.76

croft

-0.70

cember

-0.64

 Tinker

-0.63

ascript

-0.63

bay

-0.63

 Presents

-0.62

ommel

-0.61

 Events

-0.60

mania

-0.60

POSITIVE LOGITS

NRS

0.81

inqu

0.73

APS

0.71

abies

0.65

 accus

0.63

DEP

0.63

DEF

0.62

Merit

0.62

Ã©n

0.62

proc

0.61

Activations Density 0.000%

No Known Activations

This feature has no known activations.