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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    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. 85536
    Prev
    Next
    INDEX
    Explanations

    temporomandibular

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on specialized, multi‐component medical terms—especially hyphenated disorder subtypes like “tension-type headache,” “masticatory muscle,” “myofascial tissue,” etc.

    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
     Forgot
    -0.07
    a
    -0.07
    observable
    -0.06
     продаж
    -0.06
     censor
    -0.06
    ERC
    -0.06
    icture
    -0.06
     diffic
    -0.06
    _CBC
    -0.06
    unct
    -0.06
    POSITIVE LOGITS
     Kendall
    0.07
    ighton
    0.06
    _loop
    0.06
     Kv
    0.06
    lá
    0.06
     fibers
    0.06
    老师
    0.06
     Wheel
    0.06
     prend
    0.06
     fixation
    0.06
    Activations Density 0.007%

    No Known Activations