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. 130413
    Prev
    Next
    INDEX
    Explanations

    hub

    np_max-act · gemini-2.0-flash

    This neuron detects occurrences of the word “hub” (including its inflected forms and as part of names like “Hubert”).

    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
    elong
    -0.08
    leine
    -0.07
    agne
    -0.07
    jong
    -0.07
    .play
    -0.07
    aille
    -0.07
    ande
    -0.07
    .Tags
    -0.07
    ritten
    -0.06
    ompson
    -0.06
    POSITIVE LOGITS
     Hub
    0.14
     hub
    0.13
    Hub
    0.13
     hubs
    0.11
    hub
    0.09
    up
    0.08
    UP
    0.08
    ubs
    0.08
     Hut
    0.08
     ham
    0.08
    Activations Density 0.004%

    No Known Activations