Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 88765
    Prev
    Next
    INDEX
    Explanations

    database searches

    np_max-act · gemini-2.0-flash

    The neuron responds to words that describe sequence annotation steps—especially terms like “similarity,” “search,” “against,” and “databases”—i.e., references to similarity‐search–based annotation.

    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
     souls
    -0.07
     NO
    -0.06
     бо
    -0.06
     TAKE
    -0.06
    За
    -0.06
     процессе
    -0.06
     Vy
    -0.06
    樓
    -0.06
     cue
    -0.06
     scorer
    -0.06
    POSITIVE LOGITS
     halluc
    0.07
    (delete
    0.07
     foliage
    0.07
    -tab
    0.06
     coronavirus
    0.06
     Higher
    0.06
    charAt
    0.06
    :@"%@",
    0.06
     Los
    0.06
    .isDirectory
    0.06
    Activations Density 0.012%

    No Known Activations