© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 109848
    Prev
    Next
    INDEX
    Explanations

    database storage and changes

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on the special header-ID marker token (“<|start_header_id|>”), i.e. it detects those document metadata tags.

    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
    �
    -0.08
    rw
    -0.07
    어
    -0.07
    λογ
    -0.07
     Advis
    -0.07
    UnderTest
    -0.07
    dense
    -0.06
    	then
    -0.06
    /New
    -0.06
    CONT
    -0.06
    POSITIVE LOGITS
    :UITableView
    0.07
    <!--↵
    0.06
    (GLFW
    0.06
    _Vector
    0.06
    …"↵↵
    0.06
    .hasOwnProperty
    0.06
     repar
    0.06
    chodu
    0.06
     GLenum
    0.06
    .textColor
    0.06
    Activations Density 0.017%

    No Known Activations