© 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. 24-LLAMASCOPE-RES-32K
    11. 21454
    Prev
    Next
    INDEX
    Explanations

    components related to obsolescence and organizational processes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 24-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L24R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.24.hook_resid_post
    Hook Layer
    24
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    imens
    -0.18
    indow
    -0.14
    antas
    -0.14
    _MISSING
    -0.14
    à¸Īำà¸ģ
    -0.14
    cen
    -0.14
    swer
    -0.14
    adier
    -0.14
    imited
    -0.14
    nish
    -0.14
    POSITIVE LOGITS
     redundant
    0.63
     unnecessary
    0.61
     redundancy
    0.50
     excess
    0.47
     surplus
    0.45
     redund
    0.44
     unused
    0.41
     unwanted
    0.40
     useless
    0.38
     waste
    0.36
    Activations Density 0.339%

    No Known Activations