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. 130756
    Prev
    Next
    INDEX
    Explanations

    conversations

    np_max-act · gemini-2.0-flash

    The neuron activates on the placeholder tokens used for character names (e.g. “NAME_1”).

    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
     týd
    -0.08
    اا
    -0.06
     onun
    -0.06
    correo
    -0.06
     rejection
    -0.06
     valu
    -0.06
     queued
    -0.06
    ncoder
    -0.06
     imitation
    -0.06
    zeigt
    -0.06
    POSITIVE LOGITS
    ポート
    0.07
    bum
    0.07
     ns
    0.07
    continuous
    0.06
    fred
    0.06
    NST
    0.06
     standard
    0.06
    Nobody
    0.06
     yelling
    0.06
    be
    0.06
    Activations Density 0.022%

    No Known Activations