© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 43016
    Prev
    Next
    INDEX
    Explanations

    sudden change

    np_max-act · gemini-2.0-flash

    The neuron activates on language describing piecewise or stepwise changes—especially terms like “piecewise,” “step,” “wise,” “abruptly,” “breakpoints,” and “linear” that signal a function defined in segments with abrupt transitions.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     mitigation
    -0.07
     subplot
    -0.07
     panorama
    -0.07
     Yang
    -0.07
     Verification
    -0.06
     هج
    -0.06
    ircles
    -0.06
    470
    -0.06
     perennial
    -0.06
     лікування
    -0.06
    POSITIVE LOGITS
    /config
    0.07
    culator
    0.07
     discontin
    0.07
     '='
    0.06
    )paren
    0.06
    高清
    0.06
    ):-
    0.06
     경
    0.06
    	al
    0.06
     TERMIN
    0.06
    Activations Density 0.014%

    No Known Activations