INDEX

Explanations

Ct

np_max-act · gemini-2.0-flash

The neuron activates on the “Ct” token in legal citation strings (e.g. “S. Ct.”), effectively detecting court-report citation markers.

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

 pouze

-0.07

	vector

-0.06

 haus

-0.06

 Whitney

-0.06

 completionHandler

-0.06

aud

-0.06

imating

-0.06

cps

-0.05

 staffer

-0.05

 Histogram

-0.05

POSITIVE LOGITS

 meeting

0.07

CK

0.07

 Clara

0.07

?):

0.07

-security

0.07

 meio

0.07

 corrective

0.06

гар

0.06

 LIFE

0.06

ck

0.06

Activations Density 0.000%

Ct

The neuron activates on the “Ct” token in legal citation strings (e.g. “S. Ct.”), effectively detecting court-report citation markers.

No Comments

No Known Activations