© 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. 29961
    Prev
    Next
    INDEX
    Explanations

    TH

    np_max-act · gemini-2.0-flash

    The neuron activates on the ordinal suffix “TH” (as in “5TH”) in legal citation references.

    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
    oven
    -0.06
    azz
    -0.06
     Canter
    -0.06
    戴
    -0.06
     Barker
    -0.06
     bord
    -0.06
    絵
    -0.06
     misplaced
    -0.06
    _we
    -0.06
    Leg
    -0.06
    POSITIVE LOGITS
    live
    0.06
     Ph
    0.06
     Extend
    0.06
     CT
    0.06
    stice
    0.06
     {\
    0.06
     đích
    0.06
     MASTER
    0.06
     Đường
    0.06
     благод
    0.06
    Activations Density 0.001%

    No Known Activations