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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 38170
    Prev
    Next
    INDEX
    Explanations

    web content

    np_max-act · gemini-2.0-flash

    The neuron activates on the special “assistant” speaker‐label token, i.e. it detects the start of an assistant’s response.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    information about the skills and qualities necessary in professional roles, particularly in accounting and financial management.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    markers denoting the assistant role or the beginning of an assistant response in chat-formatted text.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    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
     Serialized
    -0.07
     tuyển
    -0.07
     funcs
    -0.06
     thor
    -0.06
    ',
↵
    -0.06
     fg
    -0.06
     naval
    -0.06
     Hubb
    -0.06
    {}
    -0.06
    bars
    -0.06
    POSITIVE LOGITS
     erotische
    0.07
    ティ
    0.06
     protagon
    0.06
     ankles
    0.06
    _ATTACK
    0.06
     Не
    0.06
     huis
    0.06
    phant
    0.06
     À
    0.06
     आक
    0.06
    Activations Density 0.157%

    No Known Activations