© 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. Olmo-3-1125-32B
    5. 32-RES-BATCHTOPK-131K
    7. 15649
    Prev
    Next
    INDEX
    Explanations

    going downhill or at a low point

    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
     tense
    -0.09
     Cob
    -0.09
     impass
    -0.09
     memor
    -0.09
    irie
    -0.09
     inconvenience
    -0.09
     Dunn
    -0.09
     apar
    -0.09
     incid
    -0.08
     appreh
    -0.08
    POSITIVE LOGITS
     lowest
    0.33
     Lowest
    0.32
     downward
    0.31
     nad
    0.31
     deterioration
    0.30
     downhill
    0.30
     decline
    0.29
     deterior
    0.28
     lows
    0.28
     degradation
    0.27
    Activations Density 0.340%

    No Known Activations