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

    Neuronpedia

    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. 31-LLAMASCOPE-RES-32K
    11. 9205
    Prev
    Next
    INDEX
    Explanations

    various forms of the letter "d"

    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
    u
    -0.35
    ec
    -0.27
    ed
    -0.24
    ÙĬÙĨ
    -0.24
    o
    -0.24
    र
    -0.24
    an
    -0.23
    uin
    -0.23
    upe
    -0.23
    ar
    -0.23
    POSITIVE LOGITS
    ëĵĿ
    0.16
    emon
    0.15
    iesel
    0.15
    owntown
    0.15
    etermined
    0.15
    or
    0.14
    warf
    0.14
    isease
    0.14
    etroit
    0.14
    avid
    0.14
    Activations Density 0.027%

    No Known Activations