© 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. 30847
    Prev
    Next
    INDEX
    Explanations

    -f

    np_max-act · gemini-2.0-flash

    This neuron detects subword tokens that begin with a leading hyphen (i.e., hyphen-prefixed fragments).

    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
    assembly
    -0.08
    .lb
    -0.08
     hours
    -0.07
    -plane
    -0.07
     Dark
    -0.07
    booking
    -0.07
     wij
    -0.07
     disappe
    -0.07
    zero
    -0.07
    _SEC
    -0.07
    POSITIVE LOGITS
     (_,
    0.06
    .pageX
    0.06
    .ServletException
    0.06
    이션
    0.06
    ());
↵
    0.06
    (ViewGroup
    0.06
    .modules
    0.06
    》（
    0.06
    اعية
    0.06
    .MESSAGE
    0.06
    Activations Density 0.041%

    No Known Activations