© 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. 15-LLAMASCOPE-RES-131K
    11. 898
    Prev
    Next
    INDEX
    Explanations

    punctuation and its emotional impact in text

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

    The neuron detects positive subjective evaluation or praise—tokens expressing favorable sentiment or admiration.

    oai_token-act-pair · gpt-5-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
    chter
    -0.07
    à¥įतव
    -0.06
     inv
    -0.06
     Inv
    -0.06
    QUERY
    -0.06
    æĸĹ
    -0.06
     Useful
    -0.06
     others
    -0.06
    ç±
    -0.06
     lô
    -0.06
    POSITIVE LOGITS
     even
    0.10
     plus
    0.09
     Plus
    0.08
    Plus
    0.08
    even
    0.08
    _plus
    0.08
    olith
    0.07
    zos
    0.07
    plus
    0.07
     Even
    0.07
    Activations Density 0.030%

    No Known Activations