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

illegal

-0.78

omination

-0.68

Case

-0.67

kas

-0.66

models

-0.66

äºº

-0.65

DOS

-0.65

omething

-0.65

san

-0.63

 punishments

-0.62

POSITIVE LOGITS

yip

0.82

 Nepal

0.75

Mb

0.75

opter

0.74

ema

0.71

È

0.71

 Portugal

0.70

Hoy

0.69

 Viet

0.69

 Lich

0.68

Activations Density 0.000%

No Known Activations

This feature has no known activations.