© 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
    2. Olmo-3-1125-32B
    3. 32-RES-BATCHTOPK-131K
    4. 33413
    Prev
    Next
    INDEX
    Explanations

    studies and statistics suggest

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    bcywinski/Olmo-3-32B-Base-SAE/saes_allenai_Olmo-3-1125-32B_batch_top_k/resid_post_layer_32
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     obviously
    -0.14
     IMO
    -0.13
     surely
    -0.13
     fortunately
    -0.13
     hopefully
    -0.12
     undoubtedly
    -0.12
     seems
    -0.12
     seem
    -0.11
     seeming
    -0.11
     admittedly
    -0.11
    POSITIVE LOGITS
     average
    0.14
     seven
    0.13
     twice
    0.12
     statistically
    0.12
     IQ
    0.12
     Twice
    0.11
     subconscious
    0.11
     hardest
    0.11
     avg
    0.11
     actually
    0.11
    Activations Density 0.172%

    No Known Activations