© 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. 31-LLAMASCOPE-RES-32K
    11. 6446
    Prev
    Next
    INDEX
    Explanations

    mentions of specific organizations or titles related to social media

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    PageRoute
    -0.17
    afone
    -0.16
    abi
    -0.15
    üz
    -0.14
    ardi
    -0.14
    irth
    -0.14
    ynom
    -0.14
    ionage
    -0.14
    inski
    -0.13
    adelphia
    -0.13
    POSITIVE LOGITS
     loss
    0.19
     lost
    0.17
     Loss
    0.17
    loss
    0.17
    -loss
    0.16
     outside
    0.16
     Outside
    0.16
     LOSS
    0.15
    Loss
    0.15
     losses
    0.15
    Activations Density 0.006%

    No Known Activations