INDEX

Explanations

was/were

np_max-act · gemini-2.0-flash

This neuron detects mentions of experimental or analytical methods (e.g. words like “performed,” “analysis,” “conducted,” “used”).

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

 members

-0.06

poo

-0.06

讨

-0.06

 detriment

-0.06

 Bits

-0.06

 terrific

-0.06

 practically

-0.06

 Utils

-0.06

_locals

-0.06

its

-0.06

POSITIVE LOGITS

createQueryBuilder

0.07

_ylabel

0.07

--){↵

0.07

_ARCHIVE

0.07

!";↵

0.07

>↵↵

0.07

.invoke

0.07

?:

0.07

 underwent

0.07

={()

0.06

Activations Density 0.026%

was/were

This neuron detects mentions of experimental or analytical methods (e.g. words like “performed,” “analysis,” “conducted,” “used”).

No Comments

No Known Activations