INDEX

Explanations

qualifying words

np_max-act · gemini-2.0-flash

This neuron strongly activates on adverbs—words ending in “-ly.”

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

 select

-0.07

 harassment

-0.07

 climbed

-0.07

'.↵

-0.07

 Dresses

-0.07

Discussion

-0.07

-about

-0.07

Send

-0.06

MIC

-0.06

|.↵

-0.06

POSITIVE LOGITS

 Horizon

0.06

델

0.06

 Ticaret

0.06

	Namespace

0.06

्छ

0.06

 volume

0.06

Eug

0.06

:"#

0.06

ivant

0.06

 specialize

0.06

Activations Density 0.111%

qualifying words

This neuron strongly activates on adverbs—words ending in “-ly.”

No Comments

No Known Activations