© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Qwen2.5-7B-IT
    7. Resid Post - 131k
    9. 15-RESID-POST-AA
    11. 18722
    Prev
    Next
    INDEX
    Explanations

    into

    np_max-act · gemini-2.0-flash

    The neuron detects imperative dares/commands — especially coercive or sexualized instructions, often in contexts involving minors.

    oai_token-act-pair · gpt-5-mini
    Triggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-qwen2.5-7b-instruct/resid_post_layer_15/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    נית
    -0.07
    دوا
    -0.07
     browsing
    -0.07
     involved
    -0.07
    過去
    -0.07
     fora
    -0.07
     drunk
    -0.07
     login
    -0.07
    ñana
    -0.07
    nął
    -0.06
    POSITIVE LOGITS
     peng
    0.07
     statt
    0.07
    两大
    0.07
    .Success
    0.07
    aterno
    0.07
    ainment
    0.06
     Ü
    0.06
    .NAME
    0.06
    ()>
    0.06
    ⅀
    0.06
    Activations Density 0.465%

    No Known Activations