© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 21-LLAMASCOPE-RES-32K
    11. 18322
    Prev
    Next
    INDEX
    Explanations

    discussions about identity and belonging

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.21.hook_resid_post
    Hook Layer
    21
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    aggio
    -0.17
    antt
    -0.16
    ardin
    -0.15
    arden
    -0.15
    rms
    -0.14
     Gardner
    -0.14
    ä¿Ŀè¯ģ
    -0.14
    tube
    -0.14
     mean
    -0.13
    ientes
    -0.13
    POSITIVE LOGITS
     fit
    0.71
     fits
    0.70
     Fit
    0.61
     fitting
    0.59
    fit
    0.59
     Fits
    0.58
    Fit
    0.57
    fits
    0.57
    -fit
    0.57
    _fit
    0.51
    Activations Density 0.187%

    No Known Activations