INDEX

Explanations

references to Latin-related terms such as Latin American, Latin alphabet, and Latin countries

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Alcalde

-0.59

 gallina

-0.57

 Bahía

-0.57

 Cár

-0.55

 Corazón

-0.55

 Bárbara

-0.55

 vasco

-0.55

 Asunción

-0.55

basicConfig

-0.54

 Independencia

-0.54

POSITIVE LOGITS

Lat

1.30

LAT

1.24

Lat

1.19

lat

1.14

 Latin

1.14

 latin

1.11

Latin

1.08

 LATIN

1.00

 Lati

1.00

LAT

0.96

Activations Density 0.104%

references to Latin-related terms such as Latin American, Latin alphabet, and Latin countries

No Comments

No Known Activations