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. 85905
    Prev
    Next
    INDEX
    Explanations

    forum posts

    np_max-act · gemini-2.0-flash

    This neuron detects self-introduction phrases that start with “As a(n) …” (i.e. “As an indie solo developer,” “As a robotic/​mechatronic engineer,” etc.).

    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
     روست
    -0.06
     у
    -0.06
    Yahoo
    -0.06
    لود
    -0.06
    target
    -0.06
     تول
    -0.06
     architectural
    -0.06
     Chemical
    -0.06
    recover
    -0.06
     Ish
    -0.06
    POSITIVE LOGITS
     ={↵
    0.07
    ा.↵
    0.07
     IPV
    0.07
    šla
    0.06
    _SELECTOR
    0.06
    teş
    0.06
    전에
    0.06
     mutation
    0.06
    =pk
    0.06
    İ
    0.06
    Activations Density 0.113%

    No Known Activations