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. 16890
    Prev
    Next
    INDEX
    Explanations

    descriptors related to athletic performance and player attributes

    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
    /weather
    -0.15
    /docs
    -0.15
     Strateg
    -0.14
    ibold
    -0.14
    ãģŀ
    -0.14
    ãĤ¥
    -0.14
    ixer
    -0.14
    ainment
    -0.14
    Pager
    -0.13
    odus
    -0.13
    POSITIVE LOGITS
     raw
    0.22
     fluid
    0.20
     ath
    0.19
     deceptive
    0.18
     instincts
    0.17
     natural
    0.17
     lateral
    0.17
     elite
    0.17
     quick
    0.17
    raw
    0.17
    Activations Density 0.042%

    No Known Activations