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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    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. 110026
    Prev
    Next
    INDEX
    Explanations

    punctuation

    np_max-act · gemini-2.0-flash

    The neuron is selectively responding to the special “angle-bar” tokens used as document or header delimiters (e.g. `<|begin_of_text|>`, `<|start_header_id|>`, `<|end_header_id|>`, etc.).

    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
    ctrl
    -0.06
     Tip
    -0.06
     Judgment
    -0.06
    Tw
    -0.06
    ért
    -0.06
    AutoresizingMask
    -0.06
     Universe
    -0.06
     ['/
    -0.06
     rh
    -0.06
     wash
    -0.06
    POSITIVE LOGITS
    害
    0.07
    tree
    0.07
     воды
    0.06
    _blocks
    0.06
     Countries
    0.06
    structures
    0.06
     úplně
    0.06
     restrained
    0.06
     peně
    0.06
    (goal
    0.06
    Activations Density 0.042%

    No Known Activations