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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 85357
    Prev
    Next
    INDEX
    Explanations

    comma

    np_max-act · gemini-2.0-flash

    The neuron activates on “grant‐style” verbs in software license texts (e.g. merge, publish, distribute) – in other words, it spots permission/action verbs in license boilerplate.

    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
    ></
    -0.07
     リ
    -0.06
    _given
    -0.06
    Ohio
    -0.06
    figcaption
    -0.06
     남
    -0.06
    -thread
    -0.06
    Eval
    -0.06
     کمتر
    -0.06
     decipher
    -0.06
    POSITIVE LOGITS
    gy
    0.07
     imports
    0.07
     u
    0.06
    lıkla
    0.06
    вол
    0.06
     inclus
    0.06
     выход
    0.06
     вас
    0.06
    .username
    0.06
     |--------------------------------------------------------------------------↵
    0.06
    Activations Density 0.003%

    No Known Activations