INDEX

Explanations

code

np_max-act · gemini-2.0-flash

The neuron strongly activates on numeric tokens (years, version numbers, floats, etc.) typically found in comments and docblocks.

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

OrFail

-0.06

ौद

-0.06

 перепис

-0.06

classNames

-0.06

.CompilerServices

-0.06

@PostMapping

-0.06

//---------------------------------------------------------------------------↵

-0.06

 özellikle

-0.06

 steroids

-0.06

İT

-0.06

POSITIVE LOGITS

 Moran

0.07

 signup

0.06

.login

0.06

 좋은

0.06

 месяца

0.06

 creado

0.06

 popping

0.06

 وفق

0.06

Kob

0.06

(win

0.06

Activations Density 0.008%

code

The neuron strongly activates on numeric tokens (years, version numbers, floats, etc.) typically found in comments and docblocks.

No Comments

No Known Activations