© 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. 82584
    Prev
    Next
    INDEX
    Explanations

    Question-answering

    np_max-act · gemini-2.0-flash

    This neuron detects interrogative/question constructions (tokens that appear in questions — e.g., question words and the question-mark context).

    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_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
     exact
    -0.07
    elize
    -0.07
     assigned
    -0.07
    .InvariantCulture
    -0.07
    .created
    -0.06
     latitude
    -0.06
     Bentley
    -0.06
    percent
    -0.06
    劳动力
    -0.06
     Certified
    -0.06
    POSITIVE LOGITS
    armac
    0.07
    ashing
    0.07
    лог
    0.07
    Requirement
    0.07
    .ip
    0.07
    乗
    0.07
    今年以来
    0.07
     Casual
    0.06
    行為
    0.06
    ฬ
    0.06
    Activations Density 0.189%

    No Known Activations