© 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. 108944
    Prev
    Next
    INDEX
    Explanations

    "to be" verbs

    np_max-act · gemini-2.0-flash

    The neuron is a detector for the copular verb “to be” (especially the token “is”).

    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
    즘
    -0.06
     Whit
    -0.06
    eting
    -0.06
    Hip
    -0.06
     Shel
    -0.06
     Ra
    -0.06
     wind
    -0.06
     Alone
    -0.06
    Filter
    -0.06
     Stone
    -0.06
    POSITIVE LOGITS
    )))))↵
    0.07
     Monad
    0.07
     ($
    0.07
     invers
    0.06
    .',↵
    0.06
     maxX
    0.06
     boundaries
    0.06
    (ln
    0.06
     funds
    0.06
    _div
    0.06
    Activations Density 0.100%

    No Known Activations