© 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. 6804
    Prev
    Next
    INDEX
    Explanations

    unwanted opinions

    np_max-act · gemini-2.0-flash

    This neuron detects phrases expressing people’s desire (or aversion) to hear something—for example “want to hear” or “would rather not hear.”

    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
     IMPORTANT
    -0.07
     cooks
    -0.07
    (TABLE
    -0.07
     onstage
    -0.07
    ('/')↵
    -0.07
    çon
    -0.07
    _venta
    -0.07
     clarity
    -0.06
    }},↵
    -0.06
    add
    -0.06
    POSITIVE LOGITS
     ALS
    0.06
    활동
    0.06
    Prince
    0.06
    <li
    0.06
     safer
    0.06
     pang
    0.06
     betrayed
    0.05
    西
    0.05
     OPS
    0.05
     Suffolk
    0.05
    Activations Density 0.013%

    No Known Activations