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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 19-LLAMASCOPE-RES-32K
    11. 3302
    Prev
    Next
    INDEX
    Explanations

    references to significant historical events and their implications

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 19-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L19R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.19.hook_resid_post
    Hook Layer
    19
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ONTAL
    -0.19
    InView
    -0.18
    istrovstvÃŃ
    -0.17
    ávÄĽ
    -0.15
    herits
    -0.15
    ÙĬÙĨا
    -0.15
    .LayoutStyle
    -0.14
    /google
    -0.14
    \<
    -0.14
    DisplayStyle
    -0.14
    POSITIVE LOGITS
     origin
    0.30
     magnitude
    0.28
     nature
    0.27
     fame
    0.27
     importance
    0.25
     ilk
    0.23
     Importance
    0.22
     significance
    0.21
    magnitude
    0.21
     consequence
    0.21
    Activations Density 0.305%

    No Known Activations