INDEX

Explanations

his

np_max-act · gemini-2.0-flash

This neuron fires on possessive constructions (e.g. the token “his” or words with ’s).

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

Fn

-0.07

 opacity

-0.07

.OP

-0.07

 покол

-0.07

 chrom

-0.06

FM

-0.06

_AFTER

-0.06

 railroad

-0.06

Fin

-0.06

WP

-0.06

POSITIVE LOGITS

leen

0.06

，他们

0.06

NotExist

0.06

 tooltip

0.06

Nga

0.06

щення

0.06

 下载

0.06

(schedule

0.06

สว

0.05

]")]↵

0.05

Activations Density 0.006%

his

This neuron fires on possessive constructions (e.g. the token “his” or words with ’s).

No Comments

No Known Activations

his

This neuron fires on possessive constructions (e.g. the token “his” or words with ’s).

No Comments

No Known Activations