© 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. 61926
    Prev
    Next
    INDEX
    Explanations

    conversational openings and agreements

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

    markers of assistant-style chat turns and conversational openings, especially message separators and greetings.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    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
    .scalablytyped
    -0.09
    enen
    -0.09
    (ir
    -0.09
    Yep
    -0.08
    opsy
    -0.08
    /GPL
    -0.08
    ething
    -0.08
     cheer
    -0.08
     Sc
    -0.08
    ì¢Ģ
    -0.08
    POSITIVE LOGITS
     exciting
    0.15
     Sounds
    0.14
     sounds
    0.14
    Sounds
    0.14
    sounds
    0.12
     excited
    0.11
    ounds
    0.11
    rani
    0.10
     cool
    0.10
    That
    0.10
    Activations Density 0.236%

    No Known Activations