© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 2-RES-MATRYOSHKA-DC
    7. 6070
    Prev
    Next
    INDEX
    Explanations

    references to specific, intense levels of difficulty or hardness

    oai_token-act-pair · gpt-4.1-nano
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    chanind/gemma-2-2b-batch-topk-matryoshka-saes-w-32k-l0-40/standard
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    difficult
    -1.51
     Difficult
    -1.45
    Difficult
    -1.42
     difficult
    -1.41
     difficulty
    -1.16
     dificultad
    -1.13
     schwierig
    -1.09
     Difficulty
    -1.08
     schwier
    -1.08
    impossible
    -1.07
    POSITIVE LOGITS
     hard
    1.96
     Hard
    1.63
    Hard
    1.57
    hard
    1.56
     HARD
    1.37
    HARD
    1.31
    硬
    0.88
     harde
    0.84
     hards
    0.75
     硬
    0.71
    Activations Density 0.010%

    No Known Activations