INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

sie

-0.73

raltar

-0.72

abet

-0.70

aea

-0.69

/>

-0.69

ANY

-0.68

ardless

-0.67

odor

-0.67

aly

-0.67

Asia

-0.66

POSITIVE LOGITS

 Clintons

0.72

 motivation

0.71

 killer

0.67

ãģ®éŃĶ

0.67

ItemThumbnailImage

0.65

ogene

0.64

 mentality

0.63

GD

0.62

 thickness

0.62

 velocity

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.