INDEX

Explanations

code

np_max-act · gemini-2.0-flash

The neuron activates on identifier tokens—especially variable or field names that include words like “Author” or “Title” often followed by numeric suffixes.

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

_codes

-0.06

 poč

-0.06

不会

-0.06

ipple

-0.06

笑

-0.06

zaz

-0.06

IntArray

-0.06

Jak

-0.06

 yetiş

-0.06

	EXPECT

-0.06

POSITIVE LOGITS

_archive

0.07

 войны

0.07

 grab

0.07

note

0.07

 army

0.07

 компанії

0.06

 начале

0.06

 escape

0.06

 Bios

0.06

 війни

0.06

Activations Density 0.021%

code

The neuron activates on identifier tokens—especially variable or field names that include words like “Author” or “Title” often followed by numeric suffixes.

No Comments

No Known Activations