Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · GPT-OSS BatchTopK SAEs
    3. GPT-OSS-20B
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 12954
    Prev
    Next
    INDEX
    Explanations

    tucked

    np_max-act · gemini-2.0-flash

    expressions that describe spatial containment or discreet placement, such as something being tucked or positioned within, behind, or between other objects.

    oai_token-act-pair · gpt-5Triggered by @tatsatx
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-gpt-oss-20b/resid_post_layer_11/trainer_0
    Dataset (Dashboard)
    Various
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Leb
    -0.08
    tsch
    -0.08
    .transport
    -0.08
    >({↵
    -0.07
    _CONTAINER
    -0.07
     straf
    -0.07
     pid
    -0.07
     மூ
    -0.07
    Leb
    -0.07
     engage
    -0.07
    POSITIVE LOGITS
     snug
    0.10
     tucked
    0.10
     nestled
    0.09
     pockets
    0.08
     Barnes
    0.08
     nook
    0.08
     recess
    0.08
    溪
    0.08
     extras
    0.08
    ๆ
    0.08
    Activations Density 0.014%

    No Known Activations