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. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    3. Llama3.1-8B (Base)
    4. Residual Stream
    5. 25-LLAMASCOPE-RES-32K
    6. 18097
    Prev
    Next
    INDEX
    Explanations

    sections and elements related to the structure and analysis of documents

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 25-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L25R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.25.hook_resid_post
    Hook Layer
    25
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    大åħ¨
    -0.16
    .googleapis
    -0.15
    .gdx
    -0.14
     ấy
    -0.13
    ÙĬتÙĬ
    -0.13
    ,proto
    -0.13
     Specialists
    -0.12
    RECT
    -0.12
    İK
    -0.12
    ÑĨÑĸÑİ
    -0.12
    POSITIVE LOGITS
    angelog
    0.15
    raquo
    0.14
     overview
    0.14
    AW
    0.14
    emarks
    0.14
     summary
    0.14
    onica
    0.13
    ãģĿãģ®ä»ĸ
    0.13
    erner
    0.13
    vrier
    0.13
    Activations Density 0.135%

    No Known Activations