INDEX

Explanations

language/communication

np_max-act · gemini-2.0-flash

The neuron is looking for template‐style placeholder tokens marked by double braces (e.g. “{{ user }}”).

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

-0.07

_verbose

-0.07

/Public

-0.06

-filter

-0.06

 Newark

-0.06

 obesity

-0.06

véd

-0.06

 affidavit

-0.06

opathy

-0.06

POSITIVE LOGITS

xCA

0.07

.navigator

0.07

/single

0.07

 hollow

0.06

(Channel

0.06

 Сим

0.06

.habbo

0.06

 граду

0.06

허

0.06

 그러

0.06

Activations Density 0.001%

language/communication

The neuron is looking for template‐style placeholder tokens marked by double braces (e.g. “{{ user }}”).

No Comments

No Known Activations