© 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
    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. 109422
    Prev
    Next
    INDEX
    Explanations

    scientific research

    np_max-act · gemini-2.0-flash

    The neuron fires most on scholarly‐style terms that signal applying or adapting existing methods or data—words like “applicable,” “designed,” “existing,” “from,” and “extrapolated.”

    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
     był
    -0.07
     Zot
    -0.06
     getUsername
    -0.06
     setAddress
    -0.06
     ERROR
    -0.06
    .student
    -0.06
     Seahawks
    -0.06
    ))*(
    -0.06
    幼
    -0.06
    曼
    -0.06
    POSITIVE LOGITS
    imestone
    0.07
    ίτ
    0.07
     И
    0.06
    ,,,,
    0.06
    كوم
    0.06
     Defendants
    0.06
    (mt
    0.06
     Veterans
    0.06
    ,,
    0.06
    umpy
    0.06
    Activations Density 0.017%

    No Known Activations