© 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. 81731
    Prev
    Next
    INDEX
    Explanations

    hydrogen

    np_max-act · gemini-2.0-flash
    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
     BST
    -0.08
    ジャ
    -0.08
     времен
    -0.07
    遍
    -0.07
    inox
    -0.06
     toughest
    -0.06
     יהודי
    -0.06
     Talk
    -0.06
    とい
    -0.06
     rp
    -0.06
    POSITIVE LOGITS
     similarly
    0.08
    鸢
    0.07
    POSITION
    0.07
     hydrogen
    0.07
    �
    0.07
    .SetFloat
    0.07
    用了
    0.07
    应用场景
    0.07
     _)
    0.07
     accommodate
    0.07
    Activations Density 0.006%

    No Known Activations