Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 84537
    Prev
    Next
    INDEX
    Explanations

    (

    np_max-act · gemini-2.0-flash

    The neuron is tuned to detect the opening parenthesis that introduces the year in legal citation references (e.g. “(2005)”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-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
    _er
    -0.07
     Stroke
    -0.07
    高等
    -0.06
    ];//
    -0.06
    ーデ
    -0.06
     заходів
    -0.06
     tween
    -0.06
     cynical
    -0.06
     d
    -0.06
    Overrides
    -0.06
    POSITIVE LOGITS
    carry
    0.07
    вержд
    0.06
     consulta
    0.06
    т
    0.06
     ulaş
    0.06
    	version
    0.06
    ą
    0.06
    muştur
    0.06
    ments
    0.06
     descargar
    0.06
    Activations Density 0.001%

    No Known Activations

    This feature has no known activations.