INDEX

Explanations

attire

np_max-act · gemini-2.0-flash

This neuron activates on tokens that specify quantitative measures or requirements—i.e. words denoting lengths, counts, pages, eligibility, and similar numeric specifications.

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

compan

-0.07

Bai

-0.06

agenta

-0.06

ainty

-0.06

agenda

-0.06

 Yellow

-0.06

füh

-0.06

 FileUtils

-0.06

Mic

-0.06

 slogan

-0.06

POSITIVE LOGITS

的大

0.06

ustos

0.06

_ros

0.06

 poly

0.06

hos

0.06

 visually

0.06

 limbs

0.06

{\"

0.06

 carp

0.06

_MACHINE

0.06

Activations Density 0.276%

attire

This neuron activates on tokens that specify quantitative measures or requirements—i.e. words denoting lengths, counts, pages, eligibility, and similar numeric specifications.

No Comments

No Known Activations