INDEX

Explanations

place names

np_max-act · gemini-2.0-flash

The neuron responds to tokens that are capitalized or all‐uppercase (i.e. proper names, titles, and acronyms).

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

 takeover

-0.07

 Majesty

-0.07

SENS

-0.06

)viewDidLoad

-0.06

 forests

-0.06

_project

-0.06

 uncertain

-0.06

 dobr

-0.06

_folder

-0.06

 binaries

-0.06

POSITIVE LOGITS

ุนายน

0.07

precated

0.06

eci

0.06

胎

0.06

agi

0.06

 İli

0.06

elas

0.06

 furn

0.06

طقة

0.06

주는

0.06

Activations Density 0.199%

place names

The neuron responds to tokens that are capitalized or all‐uppercase (i.e. proper names, titles, and acronyms).

No Comments

No Known Activations