INDEX

Explanations

US District Court cases

np_max-act · gemini-2.0-flash

The neuron activates on header metadata of court opinions—tokens in the docket‐and‐jurisdiction lines (e.g. court names, locations, dates, and file or case number codes).

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

scanner

-0.07

 towing

-0.07

_TEXTURE

-0.07

覺

-0.06

-address

-0.06

 Monitoring

-0.06

young

-0.06

 Terrorism

-0.06

 Procedures

-0.06

program

-0.06

POSITIVE LOGITS

Longrightarrow

0.07

(pol

0.07

(-(

0.06

/new

0.06

=localhost

0.06

ça

0.06

्ण

0.06

 buffalo

0.06

 história

0.06

농

0.06

Activations Density 0.017%

US District Court cases

The neuron activates on header metadata of court opinions—tokens in the docket‐and‐jurisdiction lines (e.g. court names, locations, dates, and file or case number codes).

No Comments

No Known Activations