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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 13
    Prev
    Next
    INDEX
    Explanations

    phrases related to missed opportunities or regrets

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    The neuron activates on expressions of something being missed or overlooked.

    oai_token-act-pair · o4-miniTriggered by @-mark

    phrases indicating someone has missed an opportunity or event.

    oai_token-act-pair · gpt-4oTriggered by @-mark

    miss

    np_max-act-logits · gpt-4oTriggered by @-mark
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    anyahu
    -0.63
    \}\\
    -0.60
    AndroidJUnit
    -0.52
     παρά
    -0.50
     ulang
    -0.49
    WebMethod
    -0.48
    ябре
    -0.47
     труда
    -0.47
    }),
    -0.46
    parsers
    -0.45
    POSITIVE LOGITS
     Missed
    0.94
     missed
    0.91
    Missed
    0.91
    issauga
    0.90
     misses
    0.88
     MISS
    0.86
    issipp
    0.84
    teps
    0.82
     MISSING
    0.79
     Misses
    0.79
    Activations Density 0.099%

    No Known Activations