© 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. 90817
    Prev
    Next
    INDEX
    Explanations

    best

    np_max-act · gemini-2.0-flash

    The neuron fires on superlative “best” (e.g. in phrases like “the best way …”). It essentially detects occurrences of the word “best.”

    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
     Morr
    -0.08
    gram
    -0.07
     replication
    -0.07
     Lar
    -0.07
     comm
    -0.07
     Mong
    -0.07
    ývá
    -0.07
     craw
    -0.07
     qr
    -0.06
     trace
    -0.06
    POSITIVE LOGITS
     best
    0.19
     Best
    0.15
     BEST
    0.15
    best
    0.14
    Best
    0.13
    BEST
    0.12
    -best
    0.10
    _best
    0.10
    st
    0.09
    	best
    0.08
    Activations Density 0.046%

    No Known Activations