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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 20-LLAMASCOPE-RES-32K
    11. 13516
    Prev
    Next
    INDEX
    Explanations

    phrases expressing obligations and requirements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 20-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L20R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    iasi
    -0.16
    _warnings
    -0.15
    :CGRect
    -0.15
    HLT
    -0.14
    ousel
    -0.13
    reib
    -0.13
    andro
    -0.13
    .tif
    -0.13
    uby
    -0.12
    ysa
    -0.12
    POSITIVE LOGITS
     must
    0.65
    must
    0.60
    å¿ħé¡»
    0.58
    Must
    0.55
     Must
    0.55
     requires
    0.54
     require
    0.54
     phải
    0.50
     MUST
    0.50
     need
    0.49
    Activations Density 0.704%

    No Known Activations