INDEX

Explanations

information/instructions

np_max-act · gemini-2.0-flash

The neuron is sensitive to call-to-action and contact language, especially second-person directives (“you,” “us,” “join,” “sign up,” and email/address tokens).

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

ix

-0.06

linewidth

-0.06

 following

-0.06

░

-0.05

ynamics

-0.05

\Block

-0.05

leted

-0.05

/<

-0.05

aghan

-0.05

opor

-0.05

POSITIVE LOGITS

 Forced

0.07

Bid

0.07

(stmt

0.07

AMP

0.07

 dash

0.06

ECM

0.06

 Final

0.06

 أمريكي

0.06

 LOWER

0.06

 Oscar

0.06

Activations Density 0.169%

information/instructions

The neuron is sensitive to call-to-action and contact language, especially second-person directives (“you,” “us,” “join,” “sign up,” and email/address tokens).

No Comments

No Known Activations