© 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. 17-LLAMASCOPE-RES-131K
    6. 3187
    Prev
    Next
    INDEX
    Explanations

    inquiries about personal identity and self-reflection

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

    The highlighted text segments typically include "the" preceding noun phrases, common assessment tools (like "CliftonStrengths"), or action phrases relating to discovering or presenting information (such as "wrapped lists", "assessment", "discovering", "comprehension was"). These often appear in educational, evaluative, or informational contexts.

    eleuther_acts_top20 · claude-3-7-sonnet-20250219Triggered by @lu-christina
    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
    avers
    -0.06
    itchens
    -0.06
    edException
    -0.06
    aleb
    -0.06
    frey
    -0.06
     safety
    -0.06
     Royal
    -0.06
    kad
    -0.05
     requ
    -0.05
    /goto
    -0.05
    POSITIVE LOGITS
     personality
    0.08
     Personality
    0.08
     quiz
    0.08
    aternity
    0.08
     score
    0.07
    _self
    0.07
    .self
    0.07
     quizzes
    0.07
    _result
    0.07
     scores
    0.07
    Activations Density 0.013%

    No Known Activations