© 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. 27087
    Prev
    Next
    INDEX
    Explanations

    symbolic punctuation marks, particularly question marks and periods, indicating inquiries and statements

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

    tokens that indicate personal emotional reactions or expressive interjections (worry, excitement, shock, or emphasis).

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

    This neuron lights up on informal, interactive bits of user comments—especially question marks and small reaction/interjection tokens (e.g. “back,” “wow,” “now?”) that signal a conversational or reactive utterance.

    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
    ersist
    -0.08
    Äįel
    -0.07
    nemonic
    -0.07
    ë³
    -0.07
    dal
    -0.07
     kê
    -0.07
     Dude
    -0.07
    à¥ĩयर
    -0.07
    elman
    -0.07
    ellas
    -0.07
    POSITIVE LOGITS
    .gdx
    0.07
    надлеж
    0.07
    olds
    0.06
    chandle
    0.06
     RouteServiceProvider
    0.06
    enderror
    0.06
     ÐļÑĢа
    0.06
    ients
    0.06
     lots
    0.06
    -quote
    0.06
    Activations Density 0.062%

    No Known Activations