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. 85474
    Prev
    Next
    INDEX
    Explanations

    numbers of teams in leagues

    np_max-act · gemini-2.0-flash

    The neuron activates on tokens representing numeric quantities (years, counts, team numbers, seasons, 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
     разд
    -0.08
    池
    -0.07
     Vaults
    -0.06
     зад
    -0.06
    тап
    -0.06
     murky
    -0.06
    statt
    -0.06
    :"-"`↵
    -0.06
    ."↵↵↵↵
    -0.06
     provisions
    -0.06
    POSITIVE LOGITS
     Electronic
    0.06
    _aug
    0.06
    ASON
    0.06
     balancing
    0.06
     september
    0.06
     FileStream
    0.06
     üye
    0.06
     finale
    0.06
     scanners
    0.06
    ccess
    0.06
    Activations Density 0.007%

    No Known Activations