INDEX

Explanations

=

np_max-act · gemini-2.0-flash

This neuron detects the Go‐generated boilerplate lines that assign imported symbols to the blank identifier—e.g. “var _ = proto.Marshal”, “var _ = fmt.Errorf”, “var _ = math.Inf”—used to suppress unused‐import errors.

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

}</

-0.07

`<

-0.07

 dearly

-0.07

 прев

-0.06

(`${

-0.06

散

-0.06

.Printf

-0.06

Ent

-0.06

 Crit

-0.06

ｨ

-0.06

POSITIVE LOGITS

駆

0.07

Widget

0.06

_IP

0.06

 Gover

0.06

 عشر

0.06

-drive

0.06

email

0.06

тот

0.06

 adamant

0.06

 prefab

0.06

Activations Density 0.002%

=

This neuron detects the Go‐generated boilerplate lines that assign imported symbols to the blank identifier—e.g. “var _ = proto.Marshal”, “var _ = fmt.Errorf”, “var _ = math.Inf”—used to suppress unused‐import errors.

No Comments

No Known Activations

=

This neuron detects the Go‐generated boilerplate lines that assign imported symbols to the blank identifier—e.g. “var _ = proto.Marshal”, “var _ = fmt.Errorf”, “var _ = math.Inf”—used to suppress unused‐import errors.

No Comments

No Known Activations