© 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. 31-LLAMASCOPE-RES-32K
    11. 14876
    Prev
    Next
    INDEX
    Explanations

    references to television shows and notable performances

    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
     Reddit
    -0.17
     Pornhub
    -0.17
     Streaming
    -0.17
     Blogs
    -0.16
     Snapchat
    -0.15
     Garmin
    -0.15
    Streamer
    -0.15
    .Stream
    -0.15
    ventus
    -0.14
     streaming
    -0.14
    POSITIVE LOGITS
     sponsor
    0.24
     Kraft
    0.23
     sponsors
    0.23
     synd
    0.22
     Mutual
    0.20
     Sponsor
    0.20
     NBC
    0.20
     sponsorship
    0.19
     Milton
    0.19
     CBS
    0.19
    Activations Density 0.072%

    No Known Activations