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

    Neuronpedia

    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. 24-LLAMASCOPE-RES-32K
    11. 18529
    Prev
    Next
    INDEX
    Explanations

    variations of the word "patch."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.24.hook_resid_post
    Hook Layer
    24
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    riers
    -0.15
    ICS
    -0.15
    á»įng
    -0.15
    สาย
    -0.14
    ssue
    -0.14
    igel
    -0.14
    aires
    -0.14
    ksen
    -0.14
    pga
    -0.14
    ornings
    -0.14
    POSITIVE LOGITS
    work
    0.36
    (patch
    0.29
     Patch
    0.27
    (es
    0.26
     patch
    0.25
    y
    0.25
    worked
    0.24
    Patch
    0.23
    ery
    0.23
    works
    0.23
    Activations Density 0.010%

    No Known Activations