© 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
    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. 581
    Prev
    Next
    INDEX
    Explanations

    watts

    np_max-act · gemini-2.0-flash

    The neuron activates on tokens representing physical measurement units (especially SI units for power and energy, such as “Watt,” “mW,” “J,” and related unit expressions).

    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
    字符
    -0.07
    .djangoproject
    -0.06
    仰
    -0.06
     gospel
    -0.06
    -most
    -0.06
     classname
    -0.06
    .bitmap
    -0.06
    _class
    -0.06
     Formatter
    -0.06
    чки
    -0.06
    POSITIVE LOGITS
     watt
    0.08
     Watt
    0.07
    ovat
    0.07
    ert
    0.07
     Watts
    0.07
    =w
    0.07
     watts
    0.07
     hil
    0.07
    \Common
    0.07
    CHEMY
    0.07
    Activations Density 0.008%

    No Known Activations