© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 17-LLAMASCOPE-RES-131K
    11. 31324
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "naturally" and its variations, indicating a focus on a natural or expected progression or characteristic in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    reau
    -0.06
    471
    -0.06
    ARRANT
    -0.06
    iniz
    -0.06
    ossible
    -0.06
    orr
    -0.06
    ugg
    -0.06
    uti
    -0.06
    ictured
    -0.06
    pk
    -0.06
    POSITIVE LOGITS
    åķ¦
    0.07
    ürlich
    0.06
    odyn
    0.06
    fully
    0.06
    clide
    0.06
    forth
    0.06
     grav
    0.06
    wich
    0.06
     Naturally
    0.06
     naturally
    0.06
    Activations Density 0.002%

    No Known Activations