© 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. 25984
    Prev
    Next
    INDEX
    Explanations

    hate/regret to inform/say

    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
     forgiveness
    -0.08
    ange
    -0.08
    cheng
    -0.08
     unacceptable
    -0.07
    -column
    -0.07
     powerless
    -0.07
    uch
    -0.07
    uppe
    -0.07
    VAS
    -0.07
     Lands
    -0.07
    POSITIVE LOGITS
     admit
    0.17
     admitting
    0.16
     disappoint
    0.16
     say
    0.16
     interrupt
    0.14
     admissions
    0.13
    say
    0.13
    .say
    0.13
    yen
    0.13
     admission
    0.13
    Activations Density 0.045%

    No Known Activations