© 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. Qwen2.5-7B-IT
    5. 20-MATRYOSHKA-65K
    7. 3358
    Prev
    Next
    INDEX
    Explanations

    Topics of discussionMethod 1 fails: MAX_ACTIVATING_TOKENS (conferences, terms, questions, illnesses, competitions, forums, s, programs, problems, things) are diverse plural nouns, not a clear specific group.Method 2 fails: TOKENS_AFTER_MAX_ACTIVATING_TOKEN (in, about, such, ,, that, they) do not share a clear pattern

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    chanind/qwen2.5-7B-it-layer-20-saes/pile/matryoshka/k-100
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ä¸ĭéĿ¢æĺ¯å°ı
    -0.25
    .Deep
    -0.22
    ActionCreators
    -0.22
     which
    -0.22
    èµĤ
    -0.19
     into
    -0.19
    åį³
    -0.19
    ä»İä¸Ń
    -0.18
    andbox
    -0.18
    usercontent
    -0.18
    POSITIVE LOGITS
    thinkable
    0.27
     beÄŁen
    0.22
    ngine
    0.21
    iferay
    0.21
    CHA
    0.21
    >,</
    0.20
    :UITableView
    0.20
     //=
    0.20
    iesta
    0.20
    %"><
    0.19
    Activations Density 1.099%

    No Known Activations