© 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 (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 55903
    Prev
    Next
    INDEX
    Explanations

    questioning/dismissing

    np_max-act · gemini-2.0-flash

    The neuron fires on language that conveys moral deviance or wrongdoing—i.e. negative, “depraved” or “twisted” descriptors.

    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
     Http
    -0.07
    ieves
    -0.07
    スカ
    -0.07
     THREAD
    -0.06
    lep
    -0.06
    _Project
    -0.06
    Regarding
    -0.06
     filming
    -0.06
    ανά
    -0.06
    巧
    -0.06
    POSITIVE LOGITS
     �
    0.07
     Structures
    0.07
     für
    0.06
     timetable
    0.06
     parties
    0.06
    tw
    0.06
     victory
    0.06
     disponible
    0.06
    (pts
    0.06
     transports
    0.06
    Activations Density 0.009%

    No Known Activations