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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 8942
    Prev
    Next
    INDEX
    Explanations

    reactions and expressions

    np_max-act · gemini-2.0-flash

    The neuron fires on little narrative action descriptions—especially stage‐direction–style or parenthetical action verbs (e.g. “smiles,” “points,” “laughed”) rather than ordinary dialogue words.

    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
    令人
    -0.06
    前
    -0.06
    接受
    -0.06
     ты
    -0.06
     sizin
    -0.06
     Kee
    -0.06
     restraining
    -0.06
    ению
    -0.06
    elsif
    -0.06
    otas
    -0.06
    POSITIVE LOGITS
     Ví
    0.07
    (userID
    0.07
     punish
    0.06
     شاخه
    0.06
     alterations
    0.06
     conveying
    0.06
    ,arg
    0.06
     scaleY
    0.06
     phản
    0.06
    _AES
    0.06
    Activations Density 0.064%

    No Known Activations