© 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. Qwen2.5-7B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 60107
    Prev
    Next
    INDEX
    Explanations

    a

    np_max-act · gemini-2.0-flash

    This neuron detects user "how-to" or instructional requests — questions asking how to do, get, or have something.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-qwen2.5-7b-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
     genotype
    -0.07
     cs
    -0.07
    ずっと
    -0.07
    のでしょう
    -0.06
    lico
    -0.06
    也只有
    -0.06
    ',(
    -0.06
     paras
    -0.06
     Ultimate
    -0.06
     playoffs
    -0.06
    POSITIVE LOGITS
    קש
    0.07
    kop
    0.07
    resentation
    0.07
     systemFontOfSize
    0.06
     Lit
    0.06
     Serving
    0.06
     dint
    0.06
     Dep
    0.06
    _swap
    0.06
     appoint
    0.06
    Activations Density 0.341%

    No Known Activations