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

nai

-0.73

psi

-0.70

 elig

-0.69

ETF

-0.67

 mathemat

-0.64

Ð¾

-0.64

ĻĤ

-0.63

gow

-0.63

owship

-0.62

TON

-0.61

POSITIVE LOGITS

ummer

0.78

iatus

0.77

 Cumm

0.68

imilar

0.67

olic

0.67

oulos

0.67

avis

0.66

cca

0.66

uca

0.65

osta

0.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.