INDEX

Explanations

numbers

np_max-act · gemini-2.0-flash

This neuron detects high-precision floating-point number tokens (e.g., coordinate values).

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

 Implements

-0.06

 evaluates

-0.06

iva

-0.06

 airplane

-0.06

 Pett

-0.06

 ож

-0.06

 Biol

-0.06

 kinds

-0.06

 plurality

-0.06

ديث

-0.06

POSITIVE LOGITS

/pm

0.07

мати

0.07

ادر

0.06

 شهری

0.06

@Entity

0.06

SAX

0.06

.forward

0.06

izedName

0.06

(sound

0.06

 قدم

0.06

Activations Density 0.006%

numbers

This neuron detects high-precision floating-point number tokens (e.g., coordinate values).

No Comments

No Known Activations