© 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. 129847
    Prev
    Next
    INDEX
    Explanations

    data/log output

    np_max-act · gemini-2.0-flash

    This neuron detects special format or metadata tokens marking system/instruction headers (e.g. the `<|…|>` tags).

    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
    013
    -0.07
    pga
    -0.07
    -running
    -0.07
     Therefore
    -0.07
     Damen
    -0.06
    plash
    -0.06
    렌
    -0.06
    _characters
    -0.06
    ObjectType
    -0.06
     ejercicio
    -0.06
    POSITIVE LOGITS
    (lon
    0.07
     broadcasts
    0.07
    _money
    0.06
    (all
    0.06
    ViewById
    0.06
    TED
    0.06
    amd
    0.06
     힘
    0.06
     Letter
    0.06
    .eulerAngles
    0.06
    Activations Density 0.012%

    No Known Activations