© 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. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 6186
    Prev
    Next
    INDEX
    Explanations

    TL

    np_max-act · gemini-2.0-flash

    The neuron specifically detects mentions of quantitative‐trait‐locus terminology (e.g. “quantitative,” “trait,” “loci,” “QTL”).

    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
     scheduling
    -0.07
    _fil
    -0.07
    robots
    -0.07
     Collections
    -0.06
     Bin
    -0.06
     rubbish
    -0.06
    "]).
    -0.06
     транспор
    -0.06
     sang
    -0.06
    еро
    -0.06
    POSITIVE LOGITS
     ruh
    0.07
    miner
    0.06
     دیگری
    0.06
     řešení
    0.06
    urgence
    0.06
    plates
    0.06
    ney
    0.06
    Queryable
    0.06
    ¶
    0.06
    (Web
    0.06
    Activations Density 0.002%

    No Known Activations