INDEX

Explanations

PREC

np_max-act · gemini-2.0-flash

This neuron is triggered by the “PREC” fragment in the header marking court opinions as “non-precedential,” i.e. it detects the “nonprecedential” label.

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

.Combine

-0.07

arshal

-0.06

卫

-0.06

efon

-0.06

LX

-0.06

สภ

-0.06

oài

-0.06

onald

-0.06

aju

-0.06

POSITIVE LOGITS

�

0.07

inc

0.07

 conducted

0.07

 distributes

0.07

 cult

0.06

 tutto

0.06

 weight

0.06

_pages

0.06

규

0.06

Activations Density 0.000%

PREC

This neuron is triggered by the “PREC” fragment in the header marking court opinions as “non-precedential,” i.e. it detects the “nonprecedential” label.

No Comments

No Known Activations