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

    Neuronpedia

    Jacobian LensNEW
    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. 38061
    Prev
    Next
    INDEX
    Explanations

    avoid

    np_max-act · gemini-2.0-flash

    The neuron fires on verbs (and related forms) that express avoidance or escape (e.g. “avoid,” “evade”).

    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
     küt
    -0.07
     kính
    -0.07
     connecting
    -0.07
     Sent
    -0.06
     кирп
    -0.06
    .concat
    -0.06
    .richTextBox
    -0.06
     mentally
    -0.06
     Pap
    -0.06
     Symphony
    -0.06
    POSITIVE LOGITS
     avoiding
    0.13
     avoid
    0.12
     avoided
    0.11
     Avoid
    0.11
     avoids
    0.10
     avoidance
    0.09
     unavoidable
    0.09
    Avoid
    0.09
    avoid
    0.08
     tránh
    0.08
    Activations Density 0.021%

    No Known Activations