© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 110436
    Prev
    Next
    INDEX
    Explanations

    other

    np_max-act · gemini-2.0-flash

    The neuron picks out the main action word in the behavior description—i.e. the present-participle (–ing) verb that names the behavior under evaluation.

    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
    (ix
    -0.08
     overlays
    -0.07
    Coll
    -0.06
    /aws
    -0.06
     onion
    -0.06
     dividing
    -0.06
    _met
    -0.06
    .View
    -0.06
    vw
    -0.06
    (Clone
    -0.06
    POSITIVE LOGITS
    がない
    0.07
    )findViewById
    0.06
     Bylo
    0.06
    )./
    0.06
     dinheiro
    0.06
     reve
    0.06
    潜
    0.06
    SocketAddress
    0.06
    ithub
    0.06
     zarar
    0.06
    Activations Density 0.008%

    No Known Activations