INDEX

Explanations

to

np_max-act · gemini-2.0-flash

The neuron activates on substantive technical terms—especially operation verbs (e.g. “writing,” “used,” “operable”) and key domain nouns (e.g. “file,” “compensator”)—that denote system components or actions.

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

 Liberation

-0.06

-dat

-0.06

 Abdullah

-0.06

 RecognitionException

-0.06

Ｒ

-0.06

Started

-0.06

.TextView

-0.06

 instances

-0.06

ritel

-0.06

 América

-0.06

POSITIVE LOGITS

 tempered

0.07

-Nazi

0.06

udded

0.06

 смот

0.06

.timeout

0.06

mast

0.06

bole

0.06

 atol

0.06

 Gül

0.06

 kalk

0.06

Activations Density 0.231%

to

The neuron activates on substantive technical terms—especially operation verbs (e.g. “writing,” “used,” “operable”) and key domain nouns (e.g. “file,” “compensator”)—that denote system components or actions.

No Comments

No Known Activations

to

The neuron activates on substantive technical terms—especially operation verbs (e.g. “writing,” “used,” “operable”) and key domain nouns (e.g. “file,” “compensator”)—that denote system components or actions.

No Comments

No Known Activations