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. 40456
    Prev
    Next
    INDEX
    Explanations

    Acronyms and abbreviations

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on short uppercase letter sequences (two‐ or three‐letter codes or acronyms) used as section or item identifiers.

    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
    dependent
    -0.07
     Deng
    -0.07
    “When
    -0.07
     TS
    -0.06
    _bed
    -0.06
    .alpha
    -0.06
     KS
    -0.06
    "When
    -0.06
     density
    -0.06
     банку
    -0.06
    POSITIVE LOGITS
    iosper
    0.07
    aisal
    0.07
     Angiospermae
    0.06
     бороть
    0.06
     puppet
    0.06
    iç
    0.06
    aryawan
    0.06
    oystick
    0.06
     Claus
    0.06
    ++++++++++++++++++++++++++++++++
    0.06
    Activations Density 0.103%

    No Known Activations