INDEX

Explanations

dialogues/conversations

np_max-act · gemini-2.0-flash

The neuron flags speaker‐label tokens that mark dialogue participants (e.g. “Person A,” “Person B,” etc.).

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 दव

-0.07

 Vulcan

-0.06

owers

-0.06

Camb

-0.06

센터

-0.06

 Angels

-0.06

 Brushes

-0.06

 improv

-0.06

LEC

-0.06

 отвеч

-0.06

POSITIVE LOGITS

Generate

0.08

(!

0.07

srv

0.07

 dzie

0.07

 kolay

0.07

.do

0.07

 зм

0.07

@$_

0.06

 ciudad

0.06

 Další

0.06

Activations Density 0.043%

dialogues/conversations

The neuron flags speaker‐label tokens that mark dialogue participants (e.g. “Person A,” “Person B,” etc.).

No Comments

No Known Activations