© 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. 128707
    Prev
    Next
    INDEX
    Explanations

    writing/conversational language

    np_max-act · gemini-2.0-flash

    This neuron activates on directive (imperative) verbs in system‐instruction prompts (e.g., “Write,” “make,” “to”).

    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
    uncia
    -0.07
     variance
    -0.06
    centage
    -0.06
     Latina
    -0.06
     Una
    -0.06
    sometimes
    -0.06
    (Build
    -0.06
     respected
    -0.06
    inp
    -0.06
     weekend
    -0.06
    POSITIVE LOGITS
    ्रस
    0.07
    _DONE
    0.07
    η
    0.07
     Männer
    0.06
     Bot
    0.06
     crank
    0.06
     слу
    0.06
    OVE
    0.06
    )prepare
    0.06
     yOffset
    0.06
    Activations Density 0.001%

    No Known Activations