© 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. 60156
    Prev
    Next
    INDEX
    Explanations

    hypothetical or counterfactual conditionals, especially “if”-clauses paired with modal “would/would have” constructions indicating unreal or imagined outcomes.

    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
    四位
    -0.08
    當地
    -0.07
     علم
    -0.07
     André
    -0.07
    阉
    -0.06
    владе
    -0.06
    CLE
    -0.06
    姊
    -0.06
     concentrate
    -0.06
    外汇
    -0.06
    POSITIVE LOGITS
    _qp
    0.08
    ㅏ
    0.07
     resh
    0.07
    eh
    0.07
     vitality
    0.07
     SH
    0.07
    //----------------------------------------------------------------------------↵
    0.07
     rip
    0.07
    -->
    0.06
    artisanlib
    0.06
    Activations Density 0.037%

    No Known Activations