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. 86106
    Prev
    Next
    INDEX
    Explanations

    Code

    np_max-act · gemini-2.0-flash

    This neuron primarily identifies floating-point number literals (numeric constants with decimal points) in the text.

    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
     đậu
    -0.06
     František
    -0.06
    onders
    -0.06
    ="<?=
    -0.06
     essays
    -0.06
     дос
    -0.06
    Saudi
    -0.06
     INITIAL
    -0.06
     Judith
    -0.06
     Dedicated
    -0.06
    POSITIVE LOGITS
    jure
    0.06
    (mask
    0.06
     yapar
    0.06
    _cap
    0.06
    _counters
    0.06
    _APPLICATION
    0.06
     Національ
    0.06
    вердж
    0.06
    ोषण
    0.06
    _bp
    0.06
    Activations Density 0.017%

    No Known Activations