© 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. 41605
    Prev
    Next
    INDEX
    Explanations

    ic/otic endings

    np_max-act · gemini-2.0-flash

    This neuron detects fragments of biomedical adjectives built from the “pro-” prefix (and its extensions), such as pro-fibrotic, pro-apoptotic, pro-inflammatory, pro-oxidant, and similar “pro-…” 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
     omez
    -0.07
    neighbors
    -0.07
     memnun
    -0.06
     był
    -0.06
    opal
    -0.06
     convinced
    -0.06
     zombies
    -0.06
     undertaken
    -0.06
    зации
    -0.06
    迅
    -0.06
    POSITIVE LOGITS
    斗
    0.07
     FormBuilder
    0.07
     outgoing
    0.06
     Dess
    0.06
     Conserv
    0.06
    (auth
    0.06
     warmer
    0.06
    Vect
    0.06
     pagan
    0.06
    (prod
    0.06
    Activations Density 0.009%

    No Known Activations