© 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. 49283
    Prev
    Next
    INDEX
    Explanations

    bot

    np_max-act · gemini-2.0-flash

    assistant responses that deliver concise, deterministic answers (e.g., direct calculations, translations, dates, or short factual replies).

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

    the start of concise, definitive assistant replies that directly present a computed or translated result.

    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
     מי
    -0.07
    .Fixed
    -0.07
    Detail
    -0.07
     noble
    -0.07
     nutrition
    -0.06
     thần
    -0.06
     truyền
    -0.06
    강
    -0.06
    .Queue
    -0.06
     chất
    -0.06
    POSITIVE LOGITS
     hton
    0.07
    xCF
    0.07
    萣
    0.07
    مشاه
    0.07
    同事们
    0.07
    phans
    0.06
     Incident
    0.06
    yet
    0.06
    问题
    0.06
    ployment
    0.06
    Activations Density 0.080%

    No Known Activations