© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 89698
    Prev
    Next
    INDEX
    Explanations

    power and rating

    np_max-act · gemini-2.0-flash

    The neuron activates on numeric technical specifications—especially electrical or mechanical measurement values (e.g. power ratings, voltages, currents, speeds) given with units.

    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
    حم
    -0.07
    sets
    -0.07
    _rad
    -0.06
    able
    -0.06
    _version
    -0.06
     дисцип
    -0.06
    Timeline
    -0.06
    ullah
    -0.06
    uddle
    -0.06
     cabins
    -0.06
    POSITIVE LOGITS
     Something
    0.07
     █
    0.07
     dereg
    0.07
    $L
    0.06
     Sized
    0.06
    ","");↵
    0.06
     pakistan
    0.06
     عدد
    0.06
     عند
    0.06
     consolid
    0.06
    Activations Density 0.041%

    No Known Activations