INDEX

Explanations

numbers

np_max-act · gemini-2.0-flash

The neuron fires on numeric version tokens (especially “13”) in model names like “vicuna-13b.”

oai_token-act-pair · o4-mini Triggered by @xinyanhu8

New Auto-Interp

Configuration

andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1

Dataset (Dashboard)

Various

Features

131,072

Data Type

float32

Hook Name

blocks.11.hook_resid_post

Architecture

standard

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Infinity

-0.06

ा�

-0.06

 honored

-0.06

登録

-0.06

 Honda

-0.06

Hen

-0.06

 admire

-0.06

ret

-0.06

και

-0.06

_Blue

-0.05

POSITIVE LOGITS

toList

0.08

Leo

0.07

ene

0.07

}}{{

0.07

 Jamal

0.07

=model

0.07

asury

0.07

特

0.06

 lastname

0.06

мар

0.06

Activations Density 0.007%

numbers

The neuron fires on numeric version tokens (especially “13”) in model names like “vicuna-13b.”

No Comments

No Known Activations