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

    Neuronpedia

    Jacobian LensNEW
    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. 63631
    Prev
    Next
    INDEX
    Explanations

    i

    np_max-act · gemini-2.0-flash

    This neuron activates for tokens that occur in assistant/generated responses (i.e., model reply turns), marking model-output text.

    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
    ワン
    -0.08
    多达
    -0.07
    ﯲ
    -0.07
     Hart
    -0.07
    orizontal
    -0.07
    cpp
    -0.06
     Presidential
    -0.06
    bach
    -0.06
     Tod
    -0.06
    นำ
    -0.06
    POSITIVE LOGITS
     אהבתי
    0.07
    感慨
    0.07
     içeris
    0.07
     בלי
    0.07
    泔
    0.07
    ၵ
    0.07
    显然是
    0.07
    .MaximizeBox
    0.07
     slash
    0.07
    _closed
    0.07
    Activations Density 0.888%

    No Known Activations