© 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. 56299
    Prev
    Next
    INDEX
    Explanations

    technical/patent documents

    np_max-act · gemini-2.0-flash

    This neuron activates on subword fragments of specialized technical or domain-specific terminology (e.g., the roots and prefixes of long compound words common in patent or manufacturing descriptions).

    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
    dojo
    -0.07
    竹
    -0.06
     пу
    -0.06
    	ImGui
    -0.06
     수
    -0.06
     CHARSET
    -0.06
    BIND
    -0.06
    Locker
    -0.06
    По
    -0.06
    leaflet
    -0.06
    POSITIVE LOGITS
    样子
    0.07
    nels
    0.07
     stanice
    0.06
     полез
    0.06
    secondary
    0.06
    (Model
    0.06
    ugging
    0.06
    (hostname
    0.06
    look
    0.06
    .static
    0.06
    Activations Density 0.072%

    No Known Activations