© 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. 74036
    Prev
    Next
    INDEX
    Explanations

    Preventing negative side effects

    np_max-act · gemini-2.0-flash

    The neuron activates on language describing reduction or degradation in performance or characteristics (e.g. deterioration, loss, diminished probability, substantial impact).

    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
     começ
    -0.07
    Strings
    -0.07
    	LOG
    -0.06
     trả
    -0.06
    (ctrl
    -0.06
    Sz
    -0.06
     ranging
    -0.06
    िलत
    -0.06
    esz
    -0.06
    imitive
    -0.06
    POSITIVE LOGITS
    阻
    0.07
     slime
    0.07
     enctype
    0.07
     Petersburg
    0.07
     따른
    0.06
    Jar
    0.06
    #undef
    0.06
     합
    0.06
    วก
    0.06
     toArray
    0.06
    Activations Density 0.037%

    No Known Activations