© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 127188
    Prev
    Next
    INDEX
    Explanations

    legal documents

    np_max-act · gemini-2.0-flash

    This neuron fires on the little “helper” words and verb forms that signal obligation, permission, or possibility (e.g. “to,” “would,” “could,” “should,” “–ing” gerunds) – in short, the modal/infinitive markers common in legal‐argument style.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Top Features by Cosine Similarity
    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
     zev
    -0.06
     gratis
    -0.06
     zap
    -0.06
     Premiership
    -0.06
    änge
    -0.06
    reh
    -0.05
     BB
    -0.05
    661
    -0.05
     Destruction
    -0.05
     bakım
    -0.05
    POSITIVE LOGITS
     adher
    0.07
     Banner
    0.07
     �
    0.07
    /big
    0.07
    ٍ
    0.06
    .Forms
    0.06
     recruitment
    0.06
     nobody
    0.06
    チャ
    0.06
     whisper
    0.06
    Activations Density 0.095%

    No Known Activations