© 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. 20094
    Prev
    Next
    INDEX
    Explanations

    dialogue turn markers

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     onBind
    -0.09
    illet
    -0.09
    oor
    -0.09
     Rena
    -0.08
    adian
    -0.08
    /control
    -0.08
     fieldValue
    -0.08
    ividual
    -0.07
    ClickListener
    -0.07
     Keyword
    -0.07
    POSITIVE LOGITS
    ¡´
    0.10
    ï¼Ĭ
    0.09
     ÃľNÄ°VERS
    0.09
    etooth
    0.09
    856
    0.08
    ¦æĥħ
    0.08
    ¶Į
    0.08
    @dynamic
    0.08
    estro
    0.08
    AndView
    0.08
    Activations Density 0.027%

    No Known Activations