© 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
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 56126
    Prev
    Next
    INDEX
    Explanations

    to this end

    np_max-act · gemini-2.0-flash

    The neuron fires on the “To…” tokens that introduce purpose‐ or goal‐statements (e.g. “To this end…”)—i.e. infinitival discourse markers signaling objectives.

    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
    ,on
    -0.07
    itte
    -0.07
    .to
    -0.07
     Stamford
    -0.07
     Server
    -0.07
     Pit
    -0.07
    _forms
    -0.07
    >To
    -0.06
    _helpers
    -0.06
     taxing
    -0.06
    POSITIVE LOGITS
    @FXML
    0.07
     которым
    0.07
     whose
    0.07
     callable
    0.06
     nemůže
    0.06
     externally
    0.06
     dostup
    0.06
    عية
    0.06
    ْع
    0.06
     ayr
    0.06
    Activations Density 0.034%

    No Known Activations