© 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. 126869
    Prev
    Next
    INDEX
    Explanations

    but

    np_max-act · gemini-2.0-flash

    The neuron activates on contrastive or adversative discourse cues—words like “but,” “however,” “also,” etc.—that signal a shift or qualification in the narrative.

    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
    ffield
    -0.07
     whiteColor
    -0.06
    eldo
    -0.06
     err
    -0.06
    _NT
    -0.06
    _SHADOW
    -0.06
    mensaje
    -0.06
    izontal
    -0.06
     Eph
    -0.06
    provided
    -0.06
    POSITIVE LOGITS
     but
    0.08
     BUT
    0.07
     ROOM
    0.06
     minimizing
    0.06
     Hungarian
    0.06
     Room
    0.06
    encing
    0.06
     situations
    0.06
    Species
    0.06
    @FindBy
    0.06
    Activations Density 0.119%

    No Known Activations