© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT
    4. Resid Post - 131k
    5. 7-RESID-POST-AA
    6. 94832
    Prev
    Next
    INDEX
    Explanations

    surprise, agreement

    np_max-act · gemini-2.0-flash

    first-person speaker tokens and short self-referential utterances (e.g., "I", "I'm", "I know", expressions of the speaker's thoughts or reactions).

    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_7/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.7.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Thi
    -0.07
    incoming
    -0.07
    straint
    -0.06
     girl
    -0.06
    ght
    -0.06
     Nine
    -0.06
    (egt
    -0.06
     آثار
    -0.06
    zeigen
    -0.06
    [num
    -0.06
    POSITIVE LOGITS
     Jakarta
    0.06
     OSI
    0.06
    ++){↵↵
    0.06
     bizarre
    0.06
    вают
    0.06
    				   
    0.06
    '][]
    0.06
    CHEDULE
    0.06
    IGATION
    0.06
     '".$_
    0.06
    Activations Density 0.108%

    No Known Activations