© 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. 13149
    Prev
    Next
    INDEX
    Explanations

    http requests

    np_max-act · gemini-2.0-flash

    This neuron detects numeric tokens—sequences of digits (often with dots) such as timestamps, IP address components, and software version numbers.

    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
    882
    -0.07
    _lbl
    -0.06
     Kam
    -0.06
    BLOCK
    -0.06
    _graph
    -0.06
    .box
    -0.06
    ipes
    -0.06
     Belt
    -0.06
    .check
    -0.06
    kad
    -0.06
    POSITIVE LOGITS
    navigation
    0.07
     český
    0.07
    aren
    0.07
     CDN
    0.07
    -helper
    0.07
     fountain
    0.07
    нім
    0.06
     antis
    0.06
     mlad
    0.06
    getDate
    0.06
    Activations Density 0.051%

    No Known Activations