INDEX

Explanations

locations or proper nouns, particularly related to specific streets, companies, or individuals

oai_token-act-pair · gpt-3.5-turbo

references to specific organizations, companies, or entities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 referen

-0.63

romeda

-0.59

 challeng

-0.56

 compr

-0.54

anwhile

-0.53

 confir

-0.52

enegger

-0.52

bom

-0.51

 Ambro

-0.51

 conflic

-0.51

POSITIVE LOGITS

 Brewing

0.60

 Holdings

0.51

 Productions

0.51

 Syndrome

0.50

Leaks

0.50

 Labs

0.50

MAX

0.49

Fest

0.49

 Photography

0.49

 Esports

0.48

Activations Density 0.818%

locations or proper nouns, particularly related to specific streets, companies, or individuals

references to specific organizations, companies, or entities

No Comments

No Known Activations