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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 89815
    Prev
    Next
    INDEX
    Explanations

    ng

    np_max-act · gemini-2.0-flash

    The neuron activates on Angular‐specific code constructs—especially form and template bindings (e.g. FormGroup/FormControl, ngModel, @ViewChild, [(ngModel)], etc.).

    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
     succès
    -0.09
     pleasures
    -0.07
     spirits
    -0.06
     windows
    -0.06
     reasonable
    -0.06
     afternoon
    -0.06
     mutation
    -0.06
     mehr
    -0.06
     demonstration
    -0.06
     cere
    -0.06
    POSITIVE LOGITS
     кін
    0.07
     ніколи
    0.06
     Url
    0.06
    -unstyled
    0.06
    oğ
    0.06
    AIT
    0.06
     측
    0.06
    .assertIn
    0.06
    xCC
    0.06
    .offsetWidth
    0.06
    Activations Density 0.011%

    No Known Activations