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

ALS

-0.78

Definition

-0.76

Proxy

-0.74

-0.71

Subject

-0.69

ById

-0.67

Anonymous

-0.67

RS

-0.67

 è£ıç

-0.66

POSITIVE LOGITS

hattan

0.77

kered

0.75

oldown

0.72

thodox

0.69

fts

0.67

 Hunts

0.66

acan

0.65

hov

0.64

inki

0.64

jiang

0.63

Activations Density 0.000%

No Known Activations

This feature has no known activations.