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. 40607
    Prev
    Next
    INDEX
    Explanations

    citations in papers

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on numerical tokens—digits, decimal numbers, and equation‐related numerals.

    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
    _DMA
    -0.08
     partic
    -0.08
    album
    -0.07
    _plane
    -0.07
     menc
    -0.06
     inv
    -0.06
    ListModel
    -0.06
     {*}
    -0.06
    (Buffer
    -0.06
     mata
    -0.06
    POSITIVE LOGITS
    -----↵
    0.06
    ystick
    0.06
     ),↵↵
    0.06
     ');
    0.06
    iversite
    0.06
    ERVE
    0.06
    )--
    0.06
     tirelessly
    0.06
     ):↵
    0.06
     fecha
    0.06
    Activations Density 0.025%

    No Known Activations