INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

noon

-0.78

 dispens

-0.67

icultural

-0.67

âĹ¼

-0.66

 âĢİ

-0.66

partisan

-0.65

profit

-0.65

 preschool

-0.64

inery

-0.60

 fulfillment

-0.60

POSITIVE LOGITS

yip

0.82

anges

0.68

Wiki

0.66

Ll

0.66

ILA

0.65

eneg

0.64

etheus

0.63

autions

0.63

azel

0.63

 Takeru

0.62

Activations Density 0.000%

No Known Activations

This feature has no known activations.