INDEX

Explanations

Excerpts of texts and emails

np_max-act · gemini-2.0-flash

This neuron detects abstract, formal or technical terms typical of specification or logic writing (e.g. “semantics,” “problems,” “solution,” “tools,” “logic”).

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

.TO

-0.07

्य

-0.06

 جنگ

-0.06

codes

-0.06

(paths

-0.06

new

-0.06

justice

-0.06

<?=

-0.06

скому

-0.05

Devices

-0.05

POSITIVE LOGITS

นคร

0.07

-tree

0.06

かい

0.06

-ins

0.06

remark

0.06

_long

0.06

 apples

0.06

 Mouse

0.06

-z

0.06

".↵↵

0.06

Activations Density 0.017%

Excerpts of texts and emails

This neuron detects abstract, formal or technical terms typical of specification or logic writing (e.g. “semantics,” “problems,” “solution,” “tools,” “logic”).

No Comments

No Known Activations