INDEX

Explanations

Actions, Efforts, Attempts

np_max-act · gemini-2.0-flash

The neuron detects strong “fix‐it” action words—verbs that signal resolving problems or improving and expediting processes (e.g. improve, resolve, expedite, process, ship).

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

결

-0.07

inces

-0.07

 méně

-0.06

oded

-0.06

 zcela

-0.06

็นพ

-0.06

 عباس

-0.06

�

-0.06

 który

-0.06

POSITIVE LOGITS

lb

0.07

 instrument

0.07

ยาน

0.07

 graves

0.06

대표

0.06

izza

0.06

 fixture

0.06

 euro

0.06

 legends

0.06

ималь

0.06

Activations Density 0.097%

Actions, Efforts, Attempts

The neuron detects strong “fix‐it” action words—verbs that signal resolving problems or improving and expediting processes (e.g. improve, resolve, expedite, process, ship).

No Comments

No Known Activations