© 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. 15-RESID-POST-AA
    11. 33075
    Prev
    Next
    INDEX
    Explanations

    articles/quantifiers

    np_max-act · gemini-2.0-flash

    the neuron detects hedging or minimizing phrases (e.g., "just a little", "a little", "it's just") that downplay or soften statements.

    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_15/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ανδ
    -0.07
    とは
    -0.07
     základě
    -0.06
    otti
    -0.06
    (Adapter
    -0.06
     Extreme
    -0.06
     messy
    -0.06
    lia
    -0.06
    (suffix
    -0.06
    conomic
    -0.06
    POSITIVE LOGITS
    анія
    0.06
    c
    0.06
    -widgets
    0.06
     blogger
    0.06
    浩
    0.06
    _absolute
    0.06
    ork
    0.06
     بص
    0.06
    ере
    0.06
    aN
    0.06
    Activations Density 0.015%

    No Known Activations