Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Gemma-2-2B
    3. 12-GEMMASCOPE-TRANSCODER-16K
    4. 12034
    Prev
    Next
    INDEX
    Explanations

    This neuron seems to be activating on technical descriptions of processes, like in software or hardware, particularly when it appears to be discussing a relationship between multiple actors in the process. It also activates on text related to informal conversation, such as might appear in internet chat.

    oai_token-act-pair · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_12/width_16k/average_l0_6
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.12.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     mutual
    -0.76
    mutual
    -0.71
    Mutual
    -0.65
     herself
    -0.65
     Mutual
    -0.61
    herself
    -0.57
     hers
    -0.54
     mutually
    -0.54
    #+#
    -0.54
    BibitemShut
    -0.53
    POSITIVE LOGITS
    AndEndTag
    0.59
    Viitteet
    0.57
     onCancelled
    0.57
    DockStyle
    0.56
    djangoproject
    0.54
    addPreferredGap
    0.50
     continuant
    0.50
     invokingState
    0.48
    GOTREF
    0.47
    DoubleQuotes
    0.46
    Activations Density 3.413%

    No Known Activations