Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 130686
    Prev
    Next
    INDEX
    Explanations

    Phone numbers and addresses

    np_max-act · gemini-2.0-flash

    This neuron detects sequences of digits (often with plus signs, parentheses, and separators) corresponding to telephone numbers or phone‐number–like strings.

    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
    BYTES
    -0.07
    wner
    -0.07
    606
    -0.07
     Terminator
    -0.06
    ییر
    -0.06
    fld
    -0.06
    TypeName
    -0.06
    ilere
    -0.06
    inth
    -0.06
    manent
    -0.06
    POSITIVE LOGITS
    	server
    0.06
     pistol
    0.06
    SAT
    0.06
    *dx
    0.06
     спе
    0.06
    	NULL
    0.06
    أت
    0.06
    ekkür
    0.06
     sensitivity
    0.05
     mehr
    0.05
    Activations Density 0.004%

    No Known Activations