Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 84364
    Prev
    Next
    INDEX
    Explanations

    to

    np_max-act · gemini-2.0-flash

    The neuron activates on “to” when it introduces an infinitive verb (i.e. the “to [verb]” construction).

    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
    'email
    -0.06
    xad
    -0.06
     occas
    -0.06
    الله
    -0.06
    Contains
    -0.06
     dat
    -0.06
    .TestTools
    -0.06
     Pentagon
    -0.06
    lož
    -0.06
     succession
    -0.06
    POSITIVE LOGITS
     –↵
    0.06
     userName
    0.06
    Paid
    0.06
     στι
    0.06
    urb
    0.06
     Tat
    0.06
    $↵
    0.06
    ).↵
    0.06
    aged
    0.06
     Trio
    0.06
    Activations Density 0.044%

    No Known Activations