INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

uras

-0.86

rand

-0.78

cffffcc

-0.76

ucket

-0.75

achine

-0.72

bilt

-0.70

boa

-0.69

reenshot

-0.67

monds

-0.67

gdala

-0.67

POSITIVE LOGITS

NCT

0.76

 Hollow

0.72

 Wilkinson

0.69

au

0.67

naires

0.66

 Rove

0.66

 Enforcement

0.64

 Cumm

0.64

orescence

0.64

 --------------------

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.