INDEX

Explanations

Dialogue

np_max-act · gemini-2.0-flash

This neuron primarily activates on the speaker‐label tokens (e.g. the “NAME_#” placeholders) in the dialogue.

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

 teamwork

-0.07

 yapmaya

-0.06

 Kết

-0.06

.pageX

-0.06

 бой

-0.06

acters

-0.06

 enforce

-0.06

 Yani

-0.06

 Sür

-0.06

.AreEqual

-0.06

POSITIVE LOGITS

ervisor

0.07

 extraordinarily

0.07

 exceedingly

0.06

社會

0.06

 conceivable

0.06

zman

0.06

key

0.06

orgt

0.06

 Malaysian

0.06

獨

0.06

Activations Density 0.036%

Dialogue

This neuron primarily activates on the speaker‐label tokens (e.g. the “NAME_#” placeholders) in the dialogue.

No Comments

No Known Activations