© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 22-GEMMASCOPE-RES-16K
    6. 9066
    Prev
    Next
    INDEX
    Explanations

    code-related operations and functions involving string manipulation and data parsing

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

    Splitting strings by commas or other separators

    np_acts-logits-general · gemini-2.0-flash

    splitting strings by comma

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Hook Layer
    22
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     jadx
    -0.71
    ]")]
    -0.60
    RefNanny
    -0.59
     सत्यापित
    -0.58
    enderror
    -0.56
     '*')
    -0.56
     ?>>
    -0.56
    Sklici
    -0.56
    fjspx
    -0.56
    loggen
    -0.55
    POSITIVE LOGITS
     separation
    1.18
     separated
    1.17
     separator
    1.16
     seper
    1.13
     Separ
    1.11
     separating
    1.11
     comma
    1.10
     Separation
    1.03
    Comma
    1.00
    sep
    1.00
    Activations Density 0.235%

    No Known Activations