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. 130464
    Prev
    Next
    INDEX
    Explanations

    ey

    np_max-act · gemini-2.0-flash

    The neuron responds to numeral tokens and numerical measurements (e.g. dates, quantities, fractions) 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
    くる
    -0.08
     seafood
    -0.07
    orizontal
    -0.06
    _ACCOUNT
    -0.06
     берем
    -0.06
     propaganda
    -0.06
    OAuth
    -0.06
     sourcing
    -0.06
     about
    -0.06
    iggins
    -0.06
    POSITIVE LOGITS
     disappointed
    0.07
     pubb
    0.07
    _wall
    0.06
    पत
    0.06
     všichni
    0.06
    ...↵
    0.06
     Please
    0.06
     boiled
    0.06
     mots
    0.06
     �
    0.06
    Activations Density 0.091%

    No Known Activations