© 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
    2. Qwen3.5-9B Base
    3. 15-QWENSCOPE-RES-64K
    4. 42041
    Prev
    Next
    INDEX
    Explanations

    used descriptions**Reasoning**:Looking at `MAX_ACTIVATING_TOKENS`, I see verbs like "used", "use", "handles", "give", "advocating".Looking at `TOKENS_AFTER_MAX_ACTIVATING_TOKEN`, I see "descriptions" appearing after "give".Looking at `TOP_ACTIVATING_TEXTS`, I see phrases like "give descriptions of e.g. perfect and pluperfect tenses".Also, "used" appears frequently in `MAX_ACTIVATING_TOKENS` and in the text samples, often followed by nouns or propositions.The phrase "used descriptions" captures a common pattern: the neuron might be related to the act of using or giving descriptions. "used descriptions" fits the 3-20 word constraint and is specific

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    NECT
    -0.09
     Walters
    -0.08
     Ø®ÙĦاÙĦÙĩا
    -0.08
    ivec
    -0.08
    滨河
    -0.08
    alth
    -0.08
    ubyte
    -0.08
    WithMany
    -0.08
    scher
    -0.08
    ALSE
    -0.08
    POSITIVE LOGITS
    “
    0.07
    ene
    0.06
     ters
    0.06
    è¯Ŀ说
    0.06
    T
    0.06
     unspecified
    0.06
    为ä¿Ŀè¯ģ
    0.06
    ust
    0.06
     T
    0.06
    åĽłåľ°
    0.05
    Activations Density 0.008%

    No Known Activations