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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 130277
    Prev
    Next
    INDEX
    Explanations

    burn

    np_max-act · gemini-2.0-flash

    The neuron activates on mentions of burn injuries (the word “burn” and related burn‐specific terms) in clinical contexts.

    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
     Icons
    -0.07
    Rob
    -0.07
    hread
    -0.06
     Habitat
    -0.06
    Worker
    -0.06
    =(-
    -0.06
    Tasks
    -0.06
    líb
    -0.06
    chnitt
    -0.06
     Meng
    -0.06
    POSITIVE LOGITS
     shortage
    0.07
    -billion
    0.07
    rowave
    0.07
     Throw
    0.06
    TH
    0.06
     chóng
    0.06
    ุลาคม
    0.06
    カテゴリ
    0.06
    AREN
    0.06
     burns
    0.06
    Activations Density 0.004%

    No Known Activations