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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 fist

-0.68

 fists

-0.66

 bargain

-0.65

choes

-0.64

 buckle

-0.64

ichick

-0.64

 Bulg

-0.63

 milit

-0.62

 wage

-0.60

bands

-0.59

POSITIVE LOGITS

 Avalon

0.82

eworthy

0.82

LCS

0.76

MW

0.70

anium

0.69

iT

0.68

ember

0.67

poon

0.66

Rated

0.65

NW

0.64

Activations Density 0.000%

No Known Activations

This feature has no known activations.