INDEX

Explanations

mathematical expressions/code

np_max-act · gemini-2.0-flash

This neuron activates on the final answer reporting—strings like “answer,” “is,” and the digits of the computed result.

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

"}}>↵

-0.07

stri

-0.06

cles

-0.06

mart

-0.06

вин

-0.06

ActivityIndicatorView

-0.06

งข

-0.06

 neglected

-0.06

AV

-0.06

 McCabe

-0.06

POSITIVE LOGITS

 رسم

0.07

 cancelled

0.06

�

0.06

δα

0.06

YG

0.06

 pís

0.06

 shell

0.06

 prostit

0.06

 Vive

0.06

 pencils

0.06

Activations Density 0.009%

mathematical expressions/code

This neuron activates on the final answer reporting—strings like “answer,” “is,” and the digits of the computed result.

No Comments

No Known Activations