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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 126696
    Prev
    Next
    INDEX
    Explanations

    IR

    np_max-act · gemini-2.0-flash

    The neuron fires on the “CIR”/“CUIT” tokens in court‐of‐appeals citations (e.g. “5TH CIR.” or “CIRCUIT RULE”).

    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
     hamstring
    -0.07
    uper
    -0.07
     fibonacci
    -0.07
     DVDs
    -0.07
     komb
    -0.06
    ccd
    -0.06
     tkinter
    -0.06
    cape
    -0.06
    Ps
    -0.06
    frared
    -0.06
    POSITIVE LOGITS
     raced
    0.06
     caregivers
    0.06
     numerous
    0.06
    385
    0.06
    �
    0.06
    :black
    0.06
     discharged
    0.06
    ;/*
    0.06
    ,key
    0.06
    /port
    0.06
    Activations Density 0.000%

    No Known Activations