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

interstitial

-0.75

 Riverside

-0.67

éŃĶ

-0.65

 stroke

-0.64

Charl

-0.64

âĶĢâĶĢ

-0.61

nec

-0.60

MI

-0.59

ILLE

-0.57

Ĥª

-0.57

POSITIVE LOGITS

urrencies

0.74

uminium

0.73

igslist

0.71

amia

0.70

=-=-=-=-

0.70

sonian

0.68

 seiz

0.67

cohol

0.66

nuts

0.66

vest

0.66

Activations Density 0.000%

No Known Activations

This feature has no known activations.