© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 22-GEMMASCOPE-TRANSCODER-16K
    7. 11339
    Prev
    Next
    INDEX
    Explanations

    words or phrases that describe change over time and/or selecting letters. This is a combination of two distinct behaviors. In the second portion of the document, neuron 4 looks for instances where letters are picked

    oai_token-act-pair · gemini-2.0-flash

    past tense verbs

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.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
     indicated
    -1.02
    indicated
    -0.96
     explained
    -0.89
     illustrated
    -0.89
     stated
    -0.88
    illustrated
    -0.84
    shown
    -0.84
     specified
    -0.83
     provided
    -0.82
     described
    -0.82
    POSITIVE LOGITS
     sailed
    0.64
     slept
    0.53
     jumped
    0.50
     walked
    0.49
     flowed
    0.47
     swam
    0.47
     wandered
    0.46
     collided
    0.46
     rained
    0.45
    Went
    0.45
    Activations Density 2.934%

    No Known Activations