INDEX

Explanations

Book sections

np_max-act · gemini-2.0-flash

The neuron activates on document structural terms and metadata labels—things like section or chapter headings, table‐of‐contents entries, and bibliography or resource headings.

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

Span

-0.07

goo

-0.07

Border

-0.07

_Invoke

-0.07

 stej

-0.06

-m

-0.06

CRT

-0.06

 dinosaur

-0.06

ants

-0.06

 stereotype

-0.06

POSITIVE LOGITS

 cheats

0.06

 }])↵

0.06

 indebted

0.06

()*

0.06

%).↵↵

0.06

LocalizedMessage

0.06

 JSImport

0.06

Furthermore

0.06

Anonymous

0.06

.AutoComplete

0.06

Activations Density 0.092%

Book sections

The neuron activates on document structural terms and metadata labels—things like section or chapter headings, table‐of‐contents entries, and bibliography or resource headings.

No Comments

No Known Activations

Book sections

The neuron activates on document structural terms and metadata labels—things like section or chapter headings, table‐of‐contents entries, and bibliography or resource headings.

No Comments

No Known Activations