INDEX

Explanations

questions and answers

np_max-act · gemini-2.0-flash

The neuron activates on user question or request phrases—especially “can you,” “could you,” “what should I do,” etc.—that ask for advice or instructions.

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

ispers

-0.07

wend

-0.06

leich

-0.06

Seat

-0.06

atör

-0.06

Cycle

-0.06

%"

-0.06

INGTON

-0.06

 comfortably

-0.06

 ascii

-0.06

POSITIVE LOGITS

 mongo

0.07

_FT

0.07

 WORK

0.06

لل

0.06

');↵↵↵↵

0.06

 глав

0.06

("----------------

0.06

 chuyện

0.06

 मल

0.06

(topic

0.06

Activations Density 0.023%

questions and answers

The neuron activates on user question or request phrases—especially “can you,” “could you,” “what should I do,” etc.—that ask for advice or instructions.

No Comments

No Known Activations

questions and answers

The neuron activates on user question or request phrases—especially “can you,” “could you,” “what should I do,” etc.—that ask for advice or instructions.

No Comments

No Known Activations