INDEX

Explanations

Encoding

np_max-act · gemini-2.0-flash

The neuron fires on segments of actual code snippets—particularly C# source with keywords like `using`, `class`, `namespace`, etc.—i.e. it detects blocks of programming code.

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

 pigment

-0.07

 GETGLOBAL

-0.07

 luaL

-0.07

Conf

-0.07

Wel

-0.06

_points

-0.06

мовір

-0.06

AnimationFrame

-0.06

Player

-0.06

-hook

-0.06

POSITIVE LOGITS

 ASIC

0.07

 προϊ

0.07

듣

0.06

HeaderInSection

0.06

 pově

0.06

 bát

0.06

TextStyle

0.06

 tentang

0.06

 tyto

0.06

Activations Density 0.002%

Encoding

The neuron fires on segments of actual code snippets—particularly C# source with keywords like `using`, `class`, `namespace`, etc.—i.e. it detects blocks of programming code.

No Comments

No Known Activations