© 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. 15-RESID-POST-AA
    11. 43767
    Prev
    Next
    INDEX
    Explanations

    neutrality or avoiding opinions

    np_max-act · gemini-2.0-flash

    requests for the assistant’s personal views or identity that trigger self-disclaimers about being an AI without opinions, feelings, or gender.

    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
    ,{"
    -0.07
     conosc
    -0.07
    .ToTable
    -0.07
     ldap
    -0.07
     detachment
    -0.07
    	status
    -0.07
    (da
    -0.07
     FUNCTION
    -0.07
     פו
    -0.07
    ,function
    -0.06
    POSITIVE LOGITS
    中期
    0.07
     ins
    0.07
    籽
    0.07
     feed
    0.07
    aras
    0.07
    Needed
    0.07
     כתוצאה
    0.06
    uli
    0.06
    碑
    0.06
    arpa
    0.06
    Activations Density 0.048%

    No Known Activations