INDEX

Explanations

code

np_max-act · gemini-2.0-flash

The neuron selectively activates on programming‐code tokens—particularly C/C++ identifiers, include paths, keywords, and numeric parameters in source or header files.

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

 Suzanne

-0.07

vec

-0.07

ğa

-0.07

 necessity

-0.07

-0.06

 Socialist

-0.06

ies

-0.06

 learns

-0.06

 분류

-0.06

Tasks

-0.06

POSITIVE LOGITS

.isDefined

0.07

*******
↵

0.06

 Trit

0.06

DEPEND

0.06

 захворю

0.06

と思

0.06

ONGO

0.06

 Toplam

0.06

 پرس

0.06

_MIC

0.06

Activations Density 0.031%

code

The neuron selectively activates on programming‐code tokens—particularly C/C++ identifiers, include paths, keywords, and numeric parameters in source or header files.

No Comments

No Known Activations

code

The neuron selectively activates on programming‐code tokens—particularly C/C++ identifiers, include paths, keywords, and numeric parameters in source or header files.

No Comments

No Known Activations