© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 17-LLAMASCOPE-RES-32K
    11. 20381
    Prev
    Next
    INDEX
    Explanations

    incidents related to attacks or confrontations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 17-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L17R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.17.hook_resid_post
    Hook Layer
    17
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Unsafe
    -0.17
     kvin
    -0.15
    asurer
    -0.15
    elper
    -0.15
     Unsafe
    -0.15
     Fé
    -0.15
    OLUTION
    -0.15
    aporation
    -0.15
    usercontent
    -0.14
    remen
    -0.14
    POSITIVE LOGITS
     attack
    0.38
     attempt
    0.35
     attempted
    0.32
     attacks
    0.31
     attempts
    0.30
     Attempt
    0.29
    attempt
    0.27
    Attempt
    0.27
     Attack
    0.27
     Attempts
    0.26
    Activations Density 0.144%

    No Known Activations