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. 130360
    Prev
    Next
    INDEX
    Explanations

    galaxies and stars

    np_max-act · gemini-2.0-flash

    The neuron detects domain‐specific astrophysics terminology—especially key celestial object and process nouns (e.g. “stars,” “galaxies,” “evolution,” “clusters”) in scientific 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
    .emp
    -0.06
    undler
    -0.06
    (condition
    -0.06
     انر
    -0.06
    .INT
    -0.06
    ে
    -0.06
     Conflict
    -0.06
     pager
    -0.06
    มากมาย
    -0.06
    「お
    -0.05
    POSITIVE LOGITS
    *>(&
    0.07
    -develop
    0.07
    idi
    0.06
     Emb
    0.06
     sexually
    0.06
     numerous
    0.06
     TLabel
    0.06
    ')."
    0.06
    ',"
    0.06
     orchestrated
    0.06
    Activations Density 0.004%

    No Known Activations