INDEX

Explanations

code

np_max-act · gemini-2.0-flash

This neuron selectively activates on C++ identifiers naming hardware or emulator interface components (e.g. class, device, and I/O interface names).

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

=all

-0.07

|=

-0.06

níku

-0.06

extAlignment

-0.06

ensen

-0.06

 cigar

-0.06

 thinker

-0.06

_it

-0.06

umor

-0.06

alyze

-0.06

POSITIVE LOGITS

 giả

0.06

 grunt

0.06

排

0.06

{$

0.06

-vis

0.05

ุญาต

0.05

gee

0.05

 conformity

0.05

DEL

0.05

‫

0.05

Activations Density 0.015%

code

This neuron selectively activates on C++ identifiers naming hardware or emulator interface components (e.g. class, device, and I/O interface names).

No Comments

No Known Activations