INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

atech

-0.75

iscal

-0.67

hra

-0.66

yip

-0.65

 certs

-0.63

alogue

-0.63

 graphs

-0.62

orts

-0.62

 exported

-0.62

 export

-0.61

POSITIVE LOGITS

 Hatt

0.75

iens

0.68

ITAL

0.68

brush

0.67

tie

0.65

 Puzzles

0.62

 Tsarnaev

0.62

raft

0.60

Constructed

0.60

 Misty

0.60

Activations Density 0.000%

No Known Activations

This feature has no known activations.