© 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. 4500
    Prev
    Next
    INDEX
    Explanations

    his

    np_max-act · gemini-2.0-flash

    high-frequency grammatical glue in simple requests—articles, prepositions, pronouns, and auxiliary/modal markers that structure “to verb” directives and everyday statements.

    oai_token-act-pair · gpt-5Triggered by @vetterc0

    function words—especially prepositions, pronouns, and auxiliary/modal verbs that frame actions or intentions.

    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
    Christmas
    -0.07
    脓
    -0.06
    ottenham
    -0.06
    >Z
    -0.06
    榮
    -0.06
    .refresh
    -0.06
    �
    -0.06
    Emergency
    -0.06
    ♑
    -0.06
    ♆
    -0.06
    POSITIVE LOGITS
     Ballet
    0.07
    쑤
    0.07
     giochi
    0.07
     modelo
    0.07
     Da
    0.07
    ignet
    0.07
     Trio
    0.07
    mite
    0.07
    Range
    0.07
    /src
    0.07
    Activations Density 0.282%

    No Known Activations