© 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. 11-RESID-POST-AA
    11. 64168
    Prev
    Next
    INDEX
    Explanations

    0

    np_max-act · gemini-2.0-flash

    the special conversation/control tokens that mark message boundaries and roles (like <|im_start|>, <|im_end|>, and similar meta tokens).

    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
    (Form
    -0.08
     Хо
    -0.08
    	item
    -0.07
    źni
    -0.07
     More
    -0.06
    ��
    -0.06
    portunity
    -0.06
    itled
    -0.06
    需
    -0.06
    飧
    -0.06
    POSITIVE LOGITS
     straight
    0.08
     muscular
    0.07
     Spreadsheet
    0.07
     interpreting
    0.07
     jwt
    0.07
     Camera
    0.07
     scrape
    0.07
    .commands
    0.07
    stable
    0.07
    RecyclerView
    0.07
    Activations Density 0.148%

    No Known Activations