© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 19722
    Prev
    Next
    INDEX
    Explanations

    selection

    np_max-act · gemini-2.0-flash

    abstract, formal noun phrases—especially “X of Y” constructions and heading-like key concepts indicating categories, quantities, or results.

    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
    金色
    -0.07
    -common
    -0.07
    _eval
    -0.06
     pued
    -0.06
    .em
    -0.06
    🤖
    -0.06
    FLOAT
    -0.06
    𝚠
    -0.06
     beforehand
    -0.06
    reet
    -0.06
    POSITIVE LOGITS
    '].'"
    0.07
    'en
    0.07
     sink
    0.07
    ankan
    0.07
    异地
    0.07
     setSearch
    0.07
    PLL
    0.07
    本地
    0.07
    人民银行
    0.06
     nerves
    0.06
    Activations Density 0.977%

    No Known Activations