© 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. 128611
    Prev
    Next
    INDEX
    Explanations

    answer, solving

    np_max-act · gemini-2.0-flash

    The neuron detects meta‐instruction language—words like “generate,” “search,” “help,” “answer,” and “results”—that signal a request to formulate or run a search query to assist in answering.

    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
     Aus
    -0.06
    GBK
    -0.06
    Aus
    -0.06
     系
    -0.06
     Netanyahu
    -0.06
    header
    -0.06
    施
    -0.06
     사람들이
    -0.06
     عکس
    -0.06
    SharedPtr
    -0.06
    POSITIVE LOGITS
    =np
    0.07
     configuring
    0.06
    ichi
    0.06
     Carb
    0.06
    reflection
    0.06
    shi
    0.06
    _CLASSES
    0.06
    quota
    0.06
     Sport
    0.06
    Ensure
    0.06
    Activations Density 0.014%

    No Known Activations