© 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. 2-LLAMASCOPE-RES-32K
    11. 7045
    Prev
    Next
    INDEX
    Explanations

    references to actions or behaviors involving significant moral or ethical implications

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 2-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L2R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    rado
    -0.17
    illing
    -0.16
    ecn
    -0.15
    lic
    -0.15
    ytic
    -0.14
    olar
    -0.14
    olding
    -0.14
    thing
    -0.14
    848
    -0.14
    arsing
    -0.14
    POSITIVE LOGITS
    upe
    0.17
    ively
    0.16
    uations
    0.15
     pcm
    0.15
    usted
    0.15
    ORA
    0.14
    -prepend
    0.14
    ouri
    0.14
    ipi
    0.14
    ега
    0.14
    Activations Density 0.011%

    No Known Activations