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

    Neuronpedia

    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. 7161
    Prev
    Next
    INDEX
    Explanations

    Question and answer pairs

    np_max-act · gemini-2.0-flash

    The neuron primarily responds to the special header tokens that mark the transition into the assistant’s reply (e.g. the `<|end_header_id|>` marker).

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

    tokens that mark or appear in assistant-generated turns (assistant role/header and reply boundaries).

    oai_token-act-pair · gpt-5-miniTriggered 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
    분
    -0.06
    ектив
    -0.06
     morality
    -0.06
     login
    -0.06
     nir
    -0.06
     placements
    -0.06
     pog
    -0.06
     nâng
    -0.06
     Highlight
    -0.06
     vans
    -0.06
    POSITIVE LOGITS
     ابراه
    0.07
    (dy
    0.06
     Affero
    0.06
    _DAYS
    0.06
    ��
    0.06
     Weak
    0.06
     불
    0.06
    IsActive
    0.06
    е
    0.06
    .Gravity
    0.06
    Activations Density 0.105%

    No Known Activations