Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 1429
    Prev
    Next
    INDEX
    Explanations

    C/C++ compilation

    np_max-act · gemini-2.0-flash

    This neuron activates on parts of platform triplets and toolchain identifiers (e.g. “arch-os-vendor” strings) in build/configuration flags.

    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
    Statistics
    -0.07
     Gorgeous
    -0.06
    erculosis
    -0.06
     subscriber
    -0.06
     badge
    -0.06
    Cards
    -0.06
     Mang
    -0.06
     Nex
    -0.06
     Exploration
    -0.06
     Facing
    -0.06
    POSITIVE LOGITS
     elic
    0.07
    -к
    0.07
     Mùa
    0.07
     jako
    0.06
     düzenlenen
    0.06
    <decltype
    0.06
    慧
    0.06
     taxable
    0.06
     amino
    0.06
    ρισ
    0.06
    Activations Density 0.003%

    No Known Activations