© 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. 54936
    Prev
    Next
    INDEX
    Explanations

    harassment with threats or false accusations

    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
     hyp
    -0.10
     rap
    -0.10
     bay
    -0.10
     f
    -0.10
     greedy
    -0.09
     timer
    -0.09
     adul
    -0.09
     rolling
    -0.09
    Brains
    -0.09
     defe
    -0.09
    POSITIVE LOGITS
     harassment
    0.25
     harass
    0.23
    har
    0.21
     harassed
    0.19
     unwanted
    0.19
     Harr
    0.18
    Har
    0.18
     stalk
    0.17
     Har
    0.17
     stalking
    0.17
    Activations Density 0.077%

    No Known Activations