© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 28-LLAMASCOPE-RES-32K
    11. 17731
    Prev
    Next
    INDEX
    Explanations

    occurrences of opinions, inquiries, and discussions involving authority or governance

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 28-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L28R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.28.hook_resid_post
    Hook Layer
    28
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    मत
    -0.16
    odb
    -0.14
    ednou
    -0.14
    Ïģιά
    -0.14
    ÅĪ
    -0.13
    ovid
    -0.13
    umo
    -0.13
    ç¿°
    -0.12
    rypton
    -0.12
    \Middleware
    -0.12
    POSITIVE LOGITS
     still
    1.40
    still
    1.28
     Still
    1.23
    Still
    1.19
     STILL
    1.16
    ä»į
    0.91
     ainda
    0.85
     ancora
    0.82
     masih
    0.82
     vẫn
    0.81
    Activations Density 1.534%

    No Known Activations