Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 41797
    Prev
    Next
    INDEX
    Explanations

    organic compounds

    np_max-act · gemini-2.0-flash

    This neuron detects mentions of volatile organic compounds (VOCs) and related terminology (e.g., “volatile organic compounds,” “VOC,” “organic compounds,” “vaporization”).

    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
    .split
    -0.07
    因此
    -0.07
    private
    -0.07
     Wanna
    -0.06
     đoạn
    -0.06
    hle
    -0.06
    _server
    -0.06
     黑
    -0.06
    _BITMAP
    -0.06
    mw
    -0.06
    POSITIVE LOGITS
     vehement
    0.06
     وق
    0.06
     VOC
    0.06
     collabor
    0.06
     حض
    0.06
     nowhere
    0.06
     حجم
    0.06
     woodworking
    0.06
     discrimin
    0.06
     قدرت
    0.06
    Activations Density 0.003%

    No Known Activations