© 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. 5870
    Prev
    Next
    INDEX
    Explanations

    statistical terms and measurements related to sensory input

    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
    اÙģÙĬØ©
    -0.10
    ardım
    -0.10
    wjgl
    -0.10
    áºŃu
    -0.09
    ë³µì§Ģ
    -0.09
     fChain
    -0.09
     +#+
    -0.09
     dostat
    -0.09
     Ấ
    -0.09
    /testify
    -0.09
    POSITIVE LOGITS
    ÂłV
    0.76
    -V
    0.73
    /V
    0.68
    ,V
    0.66
    (V
    0.66
     V
    0.65
    .V
    0.65
    _V
    0.64
    [V
    0.63
    V
    0.63
    Activations Density 3.248%

    No Known Activations