© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 125056
    Prev
    Next
    INDEX
    Explanations

    Legal and patent references

    np_max-act · gemini-2.0-flash

    This neuron strongly activates on numeric tokens—particularly the comma‐separated parts of large numbers and decimal fractions—so it’s detecting segments of multi‐part numerical values.

    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
     Elm
    -0.08
    em
    -0.08
     elm
    -0.08
     Linden
    -0.08
    yn
    -0.07
    ym
    -0.07
     Lem
    -0.07
     ylim
    -0.07
    Œ
    -0.07
     Em
    -0.07
    POSITIVE LOGITS
    300
    0.09
    900
    0.09
    800
    0.08
    500
    0.08
    (PC
    0.08
    ConnectionFactory
    0.08
    600
    0.08
    ARTH
    0.08
     ATA
    0.08
     مشار
    0.08
    Activations Density 0.141%

    No Known Activations