© 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. 15-LLAMASCOPE-RES-32K
    11. 12519
    Prev
    Next
    INDEX
    Explanations

    details and measurements related to physical structures and safety features

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 15-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L15R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ubyte
    -0.15
    cta
    -0.15
    remen
    -0.15
    .hwp
    -0.15
    опол
    -0.15
    anes
    -0.15
     Bend
    -0.14
    ounder
    -0.14
     ^{°}
    -0.14
    ektor
    -0.14
    POSITIVE LOGITS
     prec
    0.26
     safety
    0.23
     Safety
    0.23
     stability
    0.23
    Safety
    0.23
    afety
    0.21
     unstable
    0.20
     instability
    0.20
     stabil
    0.20
     Prec
    0.19
    Activations Density 0.076%

    No Known Activations