© 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 EvalsExportsinterp-engineNEWAPI 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. 125958
    Prev
    Next
    INDEX
    Explanations

    change and shifts

    np_max-act · gemini-2.0-flash

    phrases that express transitions or progress and other high-level abstract, evaluative concepts indicating significance, change, or impact.

    oai_token-act-pair · gpt-5Triggered 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
    Not in Any Lists

    No Comments

    Negative Logits
    irect
    -0.07
    riv
    -0.06
    fort
    -0.06
    inished
    -0.06
    Trou
    -0.06
     userRepository
    -0.06
    لف
    -0.06
    تأثير
    -0.06
    ritable
    -0.06
     Bolshevik
    -0.06
    POSITIVE LOGITS
     encompasses
    0.08
    !!↵↵
    0.08
     호출
    0.07
     כאן
    0.07
    (loop
    0.07
     Cena
    0.07
     recognizes
    0.06
    (xml
    0.06
     acknowledges
    0.06
    _run
    0.06
    Activations Density 5.223%

    No Known Activations