© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 36925
    Prev
    Next
    INDEX
    Explanations

    pronouns referring to people

    np_max-act · gemini-2.0-flash

    The neuron detects first-person self-references—tokens that signal the author or speaker talking about themselves or their personal experiences.

    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
    attacks
    -0.08
     Mc
    -0.07
     मल
    -0.07
     automate
    -0.07
     provisioning
    -0.07
     migr
    -0.06
    项目
    -0.06
    ilir
    -0.06
     mutable
    -0.06
     corners
    -0.06
    POSITIVE LOGITS
    итай
    0.08
    essaging
    0.07
    +$
    0.07
    CLUS
    0.06
    BOOT
    0.06
    避
    0.06
    '][$
    0.06
    UNC
    0.06
    EDITOR
    0.06
     Β
    0.06
    Activations Density 0.161%

    No Known Activations