Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    3. Llama3.1-8B
    4. Residual Stream
    5. 25-LLAMASCOPE-RES-32K
    6. 17555
    Prev
    Next
    INDEX
    Explanations

    significant plot twists or surprising outcomes in narratives

    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
    第ä¸Ģ次
    -0.15
     наÑĤ
    -0.13
    theid
    -0.13
    _WP
    -0.13
     íļ
    -0.13
     základnÃŃ
    -0.13
    pv
    -0.13
    /Input
    -0.13
     ëª
    -0.12
     StartTime
    -0.12
    POSITIVE LOGITS
     ending
    0.56
     finale
    0.55
     finish
    0.54
     conclusion
    0.54
     final
    0.52
     concluding
    0.50
     closing
    0.49
     endings
    0.47
    Conclusion
    0.47
     finishing
    0.47
    Activations Density 0.466%

    No Known Activations