© 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. 20-LLAMASCOPE-RES-32K
    11. 20383
    Prev
    Next
    INDEX
    Explanations

    references to observational equipment and vision-related terms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    acket
    -0.17
    æ¼Ĥ
    -0.17
    reeze
    -0.15
     weighting
    -0.14
     tagName
    -0.14
    usz
    -0.13
    ãģĤãĤĬãģĮãģ¨ãģĨ
    -0.13
    clado
    -0.13
     flagged
    -0.13
    à¹ĭ
    -0.13
    POSITIVE LOGITS
     lenses
    0.32
     magn
    0.31
     bin
    0.31
     lens
    0.30
     viewing
    0.29
     zoom
    0.28
     telescope
    0.27
     ey
    0.26
     scopes
    0.26
     Lens
    0.26
    Activations Density 0.079%

    No Known Activations