© 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. 16-LLAMASCOPE-RES-32K
    11. 7934
    Prev
    Next
    INDEX
    Explanations

    references to war casualties and losses

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 16-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L16R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ãĥ³ãĥĦ
    -0.15
     armed
    -0.14
    Appearance
    -0.14
     Aç
    -0.14
    ombat
    -0.14
    innamon
    -0.13
    .untracked
    -0.13
     Threat
    -0.13
    å¹
    -0.13
    pector
    -0.13
    POSITIVE LOGITS
     losses
    0.22
     toll
    0.20
     loss
    0.19
    loss
    0.19
    Loss
    0.18
     Loss
    0.17
     تÙĦÙģ
    0.16
     Attr
    0.16
    .loss
    0.15
     Casual
    0.15
    Activations Density 0.052%

    No Known Activations