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. 130321
    Prev
    Next
    INDEX
    Explanations

    stats

    np_max-act · gemini-2.0-flash

    The neuron fires on mentions of probability distributions—especially normal (Gaussian) distributions and mixture‐modeling functions in code (e.g. scipy.stats.norm, convolution of normals, “mixture model,” etc.).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    functions related to statistical distributions in Python.

    oai_token-act-pair · gpt-4o-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
    >To
    -0.07
    calar
    -0.07
    Pairs
    -0.06
     Platforms
    -0.06
     конца
    -0.06
    .banner
    -0.06
     γλώ
    -0.06
     Campo
    -0.06
    urrection
    -0.06
    ?>&
    -0.06
    POSITIVE LOGITS
    -focus
    0.06
    全面
    0.06
    weed
    0.06
    Greetings
    0.06
    -place
    0.06
     achieve
    0.06
    antan
    0.06
    _articles
    0.06
     unsus
    0.06
     LORD
    0.06
    Activations Density 0.004%

    No Known Activations