© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Llama3.3-70B-IT
    5. 50-RESID-POST-GF
    7. 8662
    Prev
    Next
    INDEX
    Explanations

    language context or instructions

    np_acts-logits-general · gemini-2.5-flash-lite

    tokens that appear immediately before special end-of-turn markers or system prompts in conversational AI training data.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @yooniel31
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With LLAMA3.3-70B-IT @ 50-resid-post-gf
    Configuration
    Goodfire/Llama-3.3-70B-Instruct-SAE-l50/Llama-3.3-70B-Instruct-SAE-l50.pt
    Prompts (Dashboard)
    10,000 prompts, 128 tokens each
    Dataset (Dashboard)
    lmsys/lmsys-chat-1m
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     unf
    -0.09
    tical
    -0.08
    appen
    -0.08
    edin
    -0.08
    rary
    -0.08
    sap
    -0.08
    bral
    -0.08
     Penn
    -0.08
     ustanov
    -0.08
    iazza
    -0.08
    POSITIVE LOGITS
     MetroFramework
    0.10
    OrCreate
    0.09
     Kak
    0.09
    rig
    0.09
     pod
    0.09
    )((((
    0.09
    धर
    0.08
    _specific
    0.08
    PLUGIN
    0.08
    WithValue
    0.08
    Activations Density 0.054%

    No Known Activations