© 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. 91549
    Prev
    Next
    INDEX
    Explanations

    sum

    np_max-act · gemini-2.0-flash

    The neuron detects meta-instructions or system-level directives telling the assistant how to format or complete its answer.

    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
     burada
    -0.07
    _end
    -0.07
    biased
    -0.07
     phối
    -0.07
    uran
    -0.06
    achte
    -0.06
    .books
    -0.06
    vail
    -0.06
    assigned
    -0.06
    ises
    -0.06
    POSITIVE LOGITS
     Kingdom
    0.07
     kab
    0.07
     blended
    0.06
     lstm
    0.06
    _TBL
    0.06
     microsoft
    0.06
    REM
    0.06
     总
    0.06
     (\<
    0.06
     chocolate
    0.06
    Activations Density 0.006%

    No Known Activations