INDEX

Explanations

suggestions and plans

np_max-act · gemini-2.0-flash

This neuron detects turn boundaries and question markers—i.e. the end‐of‐turn tokens and “?” punctuation that signal a user’s question.

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

 morality

-0.07

Cache

-0.07

[H

-0.07

 educators

-0.07

 Cheat

-0.07

={[↵

-0.06

 tuple

-0.06

<table

-0.06

anche

-0.06

 Instructor

-0.06

POSITIVE LOGITS

 повинні

0.06

{}

0.06

 ягод

0.06

 ای

0.06

आई

0.06

�

0.06

 genitals

0.06

 chảy

0.06

 دریا

0.06

jekt

0.06

Activations Density 0.075%

suggestions and plans

This neuron detects turn boundaries and question markers—i.e. the end‐of‐turn tokens and “?” punctuation that signal a user’s question.

No Comments

No Known Activations

suggestions and plans

This neuron detects turn boundaries and question markers—i.e. the end‐of‐turn tokens and “?” punctuation that signal a user’s question.

No Comments

No Known Activations