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. 90064
    Prev
    Next
    INDEX
    Explanations

    legal documents

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on blocks that look like formal court filings or docket‐style legal documents with lots of structured headings and procedural formatting.

    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
     oxidation
    -0.08
     interactions
    -0.08
     JSX
    -0.07
     apparently
    -0.07
     experiment
    -0.07
     ed
    -0.07
     interaction
    -0.07
     on
    -0.07
     Hidden
    -0.07
     fail
    -0.07
    POSITIVE LOGITS
    ิงห
    0.06
    .getenv
    0.06
    todo
    0.06
     واست
    0.06
     odst
    0.06
    179
    0.06
     tento
    0.06
    _tabs
    0.06
    lus
    0.06
    erland
    0.06
    Activations Density 0.003%

    No Known Activations