INDEX

Explanations

punctuation

np_max-act · gemini-2.0-flash

The neuron is specialized to detect Latin‐style abbreviation tokens (e.g. “i.e.”, “e.g.”) and their punctuation.

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

 Governments

-0.07

 você

-0.07

ео

-0.07

gì

-0.07

xAA

-0.06

 cruiser

-0.06

нитель

-0.06

 MEMORY

-0.06

 Beitrag

-0.06

)}}"

-0.06

POSITIVE LOGITS

 mongo

0.06

/in

0.06

Aw

0.06

>?

0.06

ービス

0.06

cut

0.06

¥

0.06

.PageSize

0.06

 suit

0.06

Activations Density 0.018%

punctuation

The neuron is specialized to detect Latin‐style abbreviation tokens (e.g. “i.e.”, “e.g.”) and their punctuation.

No Comments

No Known Activations