© 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. 128654
    Prev
    Next
    INDEX
    Explanations

    variations and dependencies

    np_max-act · gemini-2.0-flash

    The neuron activates on hedging or qualifier words and phrases (e.g. “however,” “may,” “can vary,” “depending on,” “important to note”) that introduce uncertainty or caveats.

    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
    Player
    -0.07
     mailed
    -0.06
    handlers
    -0.06
     LCS
    -0.06
     pack
    -0.06
    culos
    -0.06
    agers
    -0.06
     тих
    -0.06
    /themes
    -0.06
     rankings
    -0.06
    POSITIVE LOGITS
     submenu
    0.07
    )((
    0.06
    addContainerGap
    0.06
    (ValueError
    0.06
     kend
    0.06
    도별
    0.06
     Aust
    0.06
    دمة
    0.06
     нер
    0.06
     Dawson
    0.06
    Activations Density 0.044%

    No Known Activations