© 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. 103828
    Prev
    Next
    INDEX
    Explanations

    pronouns and verbs indicating personal involvement or actions

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

    The neuron detects first- and second-person pronouns and related conversational verb forms that indicate personal/addressing language (e.g., "I", "we", "you", "have", "had").

    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
    ewan
    -0.09
    urv
    -0.08
    zcze
    -0.08
    uhe
    -0.07
    .dy
    -0.07
    .scalablytyped
    -0.07
    ноз
    -0.07
    omnia
    -0.07
    rez
    -0.07
    istik
    -0.07
    POSITIVE LOGITS
     who
    0.09
     whom
    0.09
     whose
    0.07
     contact
    0.07
     trust
    0.07
    whose
    0.06
     tort
    0.06
     WHO
    0.06
    who
    0.06
    ica
    0.06
    Activations Density 0.114%

    No Known Activations