INDEX

Explanations

Top lists

np_max-act · gemini-2.0-flash

The neuron is activating on the meaningful “content” words (nouns, verbs, adjectives) in user‐supplied text rather than on punctuation or common function words.

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

 pint

-0.08

.getKey

-0.08

まず

-0.07

 dialect

-0.07

DAL

-0.06

шир

-0.06

UCKET

-0.06

Player

-0.06

 VOID

-0.06

ERRU

-0.06

POSITIVE LOGITS

-heavy

0.07

 ranked

0.07

-ranked

0.06

 alan

0.06

 heavy

0.06

 пер

0.06

}")]↵

0.06

 discovered

0.06

.Home

0.06

ملكة

0.06

Activations Density 0.069%

Top lists

The neuron is activating on the meaningful “content” words (nouns, verbs, adjectives) in user‐supplied text rather than on punctuation or common function words.

No Comments

No Known Activations