© 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. 7-LLAMASCOPE-RES-32K
    11. 1430
    Prev
    Next
    INDEX
    Explanations

    instances of the word "retain" and its variations, emphasizing themes of maintaining or keeping something

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 7-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L7R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.7.hook_resid_post
    Hook Layer
    7
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Negative Logits
    rey
    -0.16
    etary
    -0.16
    anut
    -0.14
    stal
    -0.14
    204
    -0.14
    åĪĹ
    -0.14
    reen
    -0.14
    ëĿ¼ìĿ¸
    -0.14
    ãģĸ
    -0.14
    leg
    -0.13
    POSITIVE LOGITS
    /rest
    0.16
    κÎŃ
    0.16
    रण
    0.15
    ilst
    0.15
    tings
    0.15
    FileStream
    0.15
    å½
    0.15
     pace
    0.15
    lage
    0.15
    neas
    0.14
    Activations Density 0.017%

    No Known Activations