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

    Neuronpedia

    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. 127046
    Prev
    Next
    INDEX
    Explanations

    went

    np_max-act · gemini-2.0-flash

    This neuron detects modal and perfect-tense verbs expressing that something “could have” been available or done earlier (e.g., “could have had access”).

    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
    燕
    -0.07
    equal
    -0.07
    _setup
    -0.07
     nghiệp
    -0.07
     Bee
    -0.07
    _minutes
    -0.07
    HH
    -0.07
    Military
    -0.06
    ximity
    -0.06
    North
    -0.06
    POSITIVE LOGITS
     iPhones
    0.07
     Ή
    0.07
    (firstName
    0.06
     ConcurrentHashMap
    0.06
    ProcAddress
    0.06
    　　　　 　 　
    0.06
    .onClick
    0.06
    *>(&
    0.06
    Ability
    0.06
    $title
    0.06
    Activations Density 0.158%

    No Known Activations