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. 40584
    Prev
    Next
    INDEX
    Explanations

    negation/absence

    np_max-act · gemini-2.0-flash

    The neuron fires on mentions of things being absent or at minimal/zero levels (e.g. “no-X,” “none,” “extremely low X”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    The neuron flags mentions of rain or rainfall (including references to low or no precipitation).

    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
     cook
    -0.06
     GU
    -0.06
     pozit
    -0.06
    aurant
    -0.06
    _Level
    -0.06
    Fat
    -0.06
     yards
    -0.06
     enthusiast
    -0.06
    IFO
    -0.06
     Mercy
    -0.06
    POSITIVE LOGITS
    vehicle
    0.07
    _rem
    0.07
    اله
    0.07
    urb
    0.06
    "=>"
    0.06
     Cyc
    0.06
    atto
    0.06
     بگ
    0.06
    AFX
    0.06
     cyc
    0.06
    Activations Density 0.091%

    No Known Activations