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

    Neuronpedia

    Jacobian LensNEW
    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. 57025
    Prev
    Next
    INDEX
    Explanations

    organizations

    np_max-act · gemini-2.0-flash

    The neuron activates on named entities—especially organization, project, or brand names (e.g. “Towards Maturity,” “e-skills UK,” “Becta,” “NASSCOM”).

    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
    .Graph
    -0.06
     KNOW
    -0.06
     ANAL
    -0.06
     ديگر
    -0.06
    INU
    -0.06
     HID
    -0.06
    .endTime
    -0.06
    _pro
    -0.06
    інки
    -0.06
     QtGui
    -0.05
    POSITIVE LOGITS
    brief
    0.07
    것
    0.07
     getItemCount
    0.07
    indexPath
    0.06
    	glog
    0.06
    :indexPath
    0.06
    dum
    0.06
    िफ
    0.06
     letter
    0.06
     sparkling
    0.06
    Activations Density 0.036%

    No Known Activations