INDEX

Explanations

Table of contents/chapters

np_max-act · gemini-2.0-flash

The neuron detects structured document navigation markers—especially numbered section or chapter headings and table‐of‐contents entries (e.g. digits and their surrounding headings).

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

 solver

-0.06

Sie

-0.06

annels

-0.06

маг

-0.06

 طبیعی

-0.06

vince

-0.06

луч

-0.06

_positive

-0.06

rolling

-0.06

 multiples

-0.05

POSITIVE LOGITS

هم

0.07

 Ethics

0.07

 anew

0.07

-serving

0.06

 wears

0.06

壊

0.06

 عمر

0.06

vědom

0.06

ΟΝ

0.06

 بیماری

0.06

Activations Density 0.015%

Table of contents/chapters

The neuron detects structured document navigation markers—especially numbered section or chapter headings and table‐of‐contents entries (e.g. digits and their surrounding headings).

No Comments

No Known Activations