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. 86025
    Prev
    Next
    INDEX
    Explanations

    Incomplete or broken things

    np_max-act · gemini-2.0-flash

    The neuron flags references to “incomplete” or “last incomplete” parts of the text—that is, phrases pointing out missing or unfinished content requiring completion.

    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
    Federal
    -0.07
     Effects
    -0.06
     korun
    -0.06
     договор
    -0.06
     tylko
    -0.06
    개의
    -0.06
     slova
    -0.06
     dvou
    -0.06
    大家
    -0.06
     Dean
    -0.06
    POSITIVE LOGITS
    .Chart
    0.07
    .minecraftforge
    0.07
     dra
    0.06
     منط
    0.06
     InkWell
    0.06
     threesome
    0.06
     (~
    0.06
    ически
    0.06
     SharedModule
    0.06
     林
    0.06
    Activations Density 0.040%

    No Known Activations