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. 130692
    Prev
    Next
    INDEX
    Explanations

    change

    np_max-act · gemini-2.0-flash

    The neuron activates on phrases describing things that change over time—i.e. mentions of “change,” “over time,” and similar temporal‐change expressions.

    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
     HACK
    -0.07
     distributions
    -0.07
     distribution
    -0.07
    Search
    -0.06
     Compact
    -0.06
     paging
    -0.06
    render
    -0.06
    사는
    -0.06
     RAD
    -0.06
     Myth
    -0.06
    POSITIVE LOGITS
    bohydr
    0.06
     amel
    0.06
    วรรณ
    0.06
    ibernate
    0.06
    erie
    0.06
     recess
    0.06
     prio
    0.06
    _Version
    0.06
    (aux
    0.06
    .Ş
    0.06
    Activations Density 0.022%

    No Known Activations