© 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. 30-LLAMASCOPE-RES-32K
    11. 9814
    Prev
    Next
    INDEX
    Explanations

    detailed evaluations of creative works, specifically focusing on performance quality and production elements in movies and games

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.30.hook_resid_post
    Hook Layer
    30
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    umberland
    -0.15
    abar
    -0.13
    á»ijng
    -0.13
     çĻº
    -0.13
    ÏĨÏħ
    -0.13
    ków
    -0.12
     бÑĥк
    -0.12
    uC
    -0.12
    akter
    -0.12
     suche
    -0.12
    POSITIVE LOGITS
     decent
    0.49
     OK
    0.40
     ok
    0.39
     okay
    0.39
     alright
    0.37
     reasonable
    0.36
     reasonably
    0.33
     adequate
    0.33
     respectable
    0.33
     fair
    0.32
    Activations Density 0.583%

    No Known Activations