INDEX

Explanations

names of individuals or official positions in a written piece

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

<bos>

-1.01

 solve

-0.55

 zdar

-0.54

 zatem

-0.54

 give

-0.54

 read

-0.53

 trzeba

-0.52

 oznacza

-0.52

 spoko

-0.52

قديم

-0.52

POSITIVE LOGITS

Juf

1.63

 alkoh

1.47

 antik

1.42

 Khart

1.42

 Keny

1.41

 Bartholo

1.40

 optik

1.40

 Abbé

1.39

Cfr

1.39

 Sarm

1.38

Activations Density 2.269%

names of individuals or official positions in a written piece

No Comments

No Known Activations