© 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. 9053
    Prev
    Next
    INDEX
    Explanations

    can

    np_max-act · gemini-2.0-flash

    The neuron activates on modal auxiliary verbs and related helping words (e.g. “can,” “could,” “try,” “to”) used in user instructions.

    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
    μέν
    -0.07
    ject
    -0.07
     Figures
    -0.06
     Learning
    -0.06
    IL
    -0.06
    headed
    -0.06
     seekers
    -0.06
    ск
    -0.06
     dramatic
    -0.06
     Skills
    -0.06
    POSITIVE LOGITS
    พน
    0.07
    ●●
    0.07
    TV
    0.07
    ',↵↵
    0.06
    にと
    0.06
     Seattle
    0.06
     ří
    0.06
    など
    0.06
    operand
    0.06
     침
    0.06
    Activations Density 0.023%

    No Known Activations