© 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
    2. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    3. Llama3.1-8B (Base)
    4. Residual Stream
    5. 15-LLAMASCOPE-RES-131K
    6. 90186
    Prev
    Next
    INDEX
    Explanations

    specific symbols, punctuation marks, and format indicators commonly used in text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    The neuron detects document-structure and formatting/markup elements (LaTeX/math constructs, section headings/labels, metadata and other non-prose formatting tokens).

    oai_token-act-pair · gpt-5-miniTriggered by @yooniel31

    The neuron is sensitive to tokens occurring in formal or technical/mathematical contexts—e.g. LaTeX commands, variables, theorem‐ or proof‐style wording, and other formulaic expressions.

    oai_token-act-pair · o4-miniTriggered by @yooniel31
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    .googlecode
    -0.07
    izza
    -0.07
     fisse
    -0.07
    -lnd
    -0.07
    nap
    -0.06
    asia
    -0.06
    ixels
    -0.06
    isan
    -0.06
    ntp
    -0.06
    isci
    -0.06
    POSITIVE LOGITS
     Copyright
    0.07
    eval
    0.07
    ²
    0.06
    761
    0.06
    ¥
    0.06
     Cookies
    0.06
    ç³
    0.06
    866
    0.06
    _priv
    0.06
    eah
    0.06
    Activations Density 0.318%

    No Known Activations