© 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. 8772
    Prev
    Next
    INDEX
    Explanations

    Qualifying statements

    np_max-act · gemini-2.0-flash

    The neuron activates on hedge or disclaimer phrases that introduce caveats or extra context (e.g. “in case that matters”).

    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
     TextArea
    -0.08
    portrait
    -0.07
    _hresult
    -0.07
    ThanOr
    -0.07
     مشکلات
    -0.07
     Heater
    -0.07
    .getTable
    -0.07
    .setUser
    -0.06
    imbus
    -0.06
     Skeleton
    -0.06
    POSITIVE LOGITS
     ');
    0.06
     enlist
    0.06
    -rest
    0.06
    怎
    0.06
    );
    0.06
     Kerala
    0.06
    .epsilon
    0.06
    ６
    0.06
    TAIL
    0.06
     abroad
    0.06
    Activations Density 0.090%

    No Known Activations