INDEX

Explanations

Code/programming context

np_max-act · gemini-2.0-flash

This neuron detects occurrences of guidance phrased as “should not use flags” (i.e. advice about things one shouldn’t flag).

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

 Saving

-0.07

 Explosion

-0.07

箱

-0.06

 plant

-0.06

riages

-0.06

 plates

-0.06

.android

-0.06

 telephone

-0.06

 Reeves

-0.06

 days

-0.06

POSITIVE LOGITS

.assertIsNot

0.07

�

0.06

.Min

0.06

 preocup

0.06

จน

0.06

.goto

0.06

ucion

0.06

 što

0.06

 wallpaper

0.06

 مشارکت

0.06

Activations Density 0.110%

Code/programming context

This neuron detects occurrences of guidance phrased as “should not use flags” (i.e. advice about things one shouldn’t flag).

No Comments

No Known Activations