© 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. 127070
    Prev
    Next
    INDEX
    Explanations

    punctuation and code

    np_max-act · gemini-2.0-flash

    This neuron detects tokens involved in defining or assigning the assistant’s persona or role (e.g. “NAME_1,” “author,” and similar meta‐instruction placeholders).

    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
     zag
    -0.07
    mnop
    -0.06
    サー
    -0.06
    -but
    -0.06
     하지만
    -0.06
    alah
    -0.06
    ilos
    -0.06
    kan
    -0.06
    анные
    -0.06
    �
    -0.06
    POSITIVE LOGITS
     pineapple
    0.08
     incidence
    0.07
     Braun
    0.06
     BEGIN
    0.06
    Information
    0.06
     charity
    0.06
     calorie
    0.06
    ategorized
    0.06
     ček
    0.06
     univerz
    0.06
    Activations Density 0.001%

    No Known Activations