INDEX

Explanations

advice or consultation

np_max-act · gemini-2.0-flash

This neuron activates on advisory or disclaimer language—sentences where the model recommends checking sources, consulting experts, or otherwise qualifies its answer.

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

 cena

-0.06

 phone

-0.06

Vs

-0.06

ذ

-0.06

ARRAY

-0.06

coc

-0.06

 वस

-0.06

 childs

-0.06

 Bruce

-0.06

POSITIVE LOGITS

 bölge

0.07

 payable

0.07

原

0.07

0.06

(rect

0.06

`"]↵

0.06

Industry

0.06

 smelled

0.06

(gl

0.06

xEC

0.06

Activations Density 0.069%

advice or consultation

This neuron activates on advisory or disclaimer language—sentences where the model recommends checking sources, consulting experts, or otherwise qualifies its answer.

No Comments

No Known Activations

advice or consultation

This neuron activates on advisory or disclaimer language—sentences where the model recommends checking sources, consulting experts, or otherwise qualifies its answer.

No Comments

No Known Activations