© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 32593
    Prev
    Next
    INDEX
    Explanations

    references to methods related to the development and efficiency of chat systems and information retrieval tasks

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Negative Logits
    Stamped
    -0.15
     Bundy
    -0.15
    ÑĢÑĥж
    -0.15
    oft
    -0.14
    toolbox
    -0.14
     ActionTypes
    -0.14
    éļĨ
    -0.13
     ApplicationException
    -0.13
    ViewItem
    -0.13
    TTY
    -0.13
    POSITIVE LOGITS
     instead
    0.18
     unlike
    0.18
     inspired
    0.18
     existing
    0.17
    Unlike
    0.15
     Unlike
    0.15
     Instead
    0.15
    ohn
    0.15
     Inspired
    0.15
     Graves
    0.15
    Activations Density 0.052%

    No Known Activations