© 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. 31-LLAMASCOPE-RES-32K
    11. 9752
    Prev
    Next
    INDEX
    Explanations

    unique identifiers or special symbols within a text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.1-8B @ 31-llamascope-res-32k
    Configuration
    fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L31R-8x
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    Features
    32,768
    Data Type
    bfloat16
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     subur
    -0.14
     Erotische
    -0.13
     seksi
    -0.12
    .Style
    -0.11
     erotik
    -0.11
     erotica
    -0.11
     alternatives
    -0.11
     elimin
    -0.11
     Delete
    -0.11
     explos
    -0.11
    POSITIVE LOGITS
     honour
    0.29
     commem
    0.28
     commemor
    0.27
     honoured
    0.27
     honor
    0.26
     honoring
    0.25
     memorial
    0.25
     tribute
    0.24
     honored
    0.23
     Honour
    0.23
    Activations Density 0.016%

    No Known Activations