© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 7373
    Prev
    Next
    INDEX
    Explanations

    ,target

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on hyphenated compound modifiers (e.g. “second-largest,” “2-in-1,” “360-degree”), indicating that it is detecting hyphenated descriptive terms.

    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
           ↵
    -0.07
    斯
    -0.06
    ignite
    -0.06
     Vanderbilt
    -0.06
    ROS
    -0.06
     slic
    -0.06
     ever
    -0.06
    eniable
    -0.06
    심
    -0.06
     gì
    -0.06
    POSITIVE LOGITS
     đạt
    0.07
     товари
    0.07
    Пол
    0.07
     fellow
    0.06
     incompetence
    0.06
     Hos
    0.06
    IRROR
    0.06
    per
    0.06
     single
    0.06
    .setProperty
    0.06
    Activations Density 0.554%

    No Known Activations