Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 125912
    Prev
    Next
    INDEX
    Explanations

    )

    np_max-act · gemini-2.0-flash

    This neuron primarily activates on common English function words and short verbs (“as,” “to,” “from,” “must,” etc.) found in the license‐style comment prose rather than on code identifiers.

    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
    .Views
    -0.07
    .win
    -0.06
    vä
    -0.06
    区域
    -0.06
    .getLabel
    -0.06
     zelf
    -0.06
     เว
    -0.06
    urator
    -0.06
    USB
    -0.06
    زان
    -0.06
    POSITIVE LOGITS
    Envelope
    0.07
     Genome
    0.07
     Published
    0.07
     ساز
    0.07
     Newman
    0.07
    ;↵
    0.06
     німець
    0.06
     Vers
    0.06
     trò
    0.06
     Seymour
    0.06
    Activations Density 0.001%

    No Known Activations