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

 Flavoring

-1.15

 Landing

-0.82

Magikarp

-0.77

èĢħ

-0.77

esome

-0.75

Pad

-0.73

ppa

-0.72

issance

-0.71

âĸĳ

-0.70

Fey

-0.70

POSITIVE LOGITS

 incrim

0.79

 debunked

0.79

 unsub

0.75

 ousted

0.74

 voic

0.71

 remed

0.70

 recons

0.69

 conting

0.67

 surv

0.67

 contradict

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.