© 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 (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 85694
    Prev
    Next
    INDEX
    Explanations

    cavalry

    np_max-act · gemini-2.0-flash

    This neuron activates on mentions of cavalry troop types (e.g., cavalry, uhlans, dragoons, cuirassiers).

    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
    wake
    -0.06
     nuest
    -0.06
     isol
    -0.06
     burden
    -0.06
    온
    -0.06
     Whites
    -0.06
    Streams
    -0.06
     show
    -0.06
     Bird
    -0.06
    (filter
    -0.06
    POSITIVE LOGITS
     cavalry
    0.11
     bev
    0.07
    _dec
    0.07
    кувати
    0.06
    .fe
    0.06
    /v
    0.06
    骑
    0.06
     출연
    0.06
    AILABLE
    0.06
    .camel
    0.06
    Activations Density 0.004%

    No Known Activations