© 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 (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 110096
    Prev
    Next
    INDEX
    Explanations

    {

    np_max-act · gemini-2.0-flash

    This neuron responds to LaTeX/math‐mode formatting tokens (e.g. “\text”, “{”, “$”), i.e. the markup around mathematical expressions.

    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
    ング
    -0.07
    Todd
    -0.07
    نت
    -0.07
     dần
    -0.06
    ador
    -0.06
     Rudy
    -0.06
    pageNumber
    -0.06
     तह
    -0.06
    ,都
    -0.06
    "+↵
    -0.06
    POSITIVE LOGITS
     ttk
    0.06
    �
    0.06
    ->{
    0.06
    'email
    0.06
    lanma
    0.06
     Json
    0.06
     гар
    0.06
     eco
    0.06
    	sound
    0.06
     شع
    0.06
    Activations Density 0.004%

    No Known Activations