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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 31-GEMMASCOPE-RES-16K
    6. 13257
    Prev
    Next
    INDEX
    Explanations

    variations of the word "hard"

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

    the word "hard" paired with hyphens and following descriptive words.

    oai_token-act-pair · claude-4-5-haikuTriggered by @emiglarou
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 31-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_31/width_16k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     censo
    -0.27
    yfik
    -0.26
    readdir
    -0.25
     Trá
    -0.25
     IFTT
    -0.24
    そ
    -0.23
     Jambi
    -0.23
     RVA
    -0.22
    Appointment
    -0.22
     recevez
    -0.22
    POSITIVE LOGITS
     Hard
    1.25
    Hard
    1.24
     hard
    1.16
     HARD
    1.11
    hard
    1.10
    HARD
    1.09
     harde
    0.96
     BoxFit
    0.93
     harder
    0.90
     hård
    0.85
    Activations Density 0.090%

    No Known Activations