INDEX

Explanations

numbers

np_max-act · gemini-2.0-flash

This neuron activates on legal citation markers, especially the volume and page numbers (and associated reporter abbreviations) of court case references.

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Tank

-0.07

err

-0.06

hits

-0.06

stown

-0.06

pred

-0.06

 inplace

-0.06

(conn

-0.06

pkt

-0.06

anes

-0.05

POSITIVE LOGITS

.rdf

0.07

 Africa

0.07

 precedence

0.06

寶

0.06

HorizontalAlignment

0.06

Geo

0.06

 آز

0.06

 Contrast

0.06

baar

0.06

ahoma

0.06

Activations Density 0.005%

numbers

This neuron activates on legal citation markers, especially the volume and page numbers (and associated reporter abbreviations) of court case references.

No Comments

No Known Activations