© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 31-LLAMASCOPE-RES-32K
    11. 16542
    Prev
    Next
    INDEX
    Explanations

    information related to fees and charges associated with services

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ï¼Įä¸įè¿ĩ
    -0.16
    ãģıãģł
    -0.16
    ï¼Įåį´
    -0.15
    ãģªãģ®
    -0.15
    inki
    -0.15
    ãĤĪãģŃ
    -0.15
    orgot
    -0.15
     zoekt
    -0.15
    ãģ¨æĢĿãģĨ
    -0.14
    ãģłãģª
    -0.14
    POSITIVE LOGITS
    ï¼Į
    0.31
    ï¼İ
    0.29
    ï¼Ī
    0.28
     ï¼Į
    0.28
     ï¼Ī
    0.28
    ï¼Ľ
    0.26
    ï¼ļ
    0.26
    ï¼ļ↵
    0.24
    ï¼
    0.24
     ï¼ļ
    0.24
    Activations Density 0.119%

    No Known Activations