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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 5690
    Prev
    Next
    INDEX
    Explanations

    The neuron is looking for the name "Duarte"

    oai_token-act-pair · gpt-3.5-turbo

    references to a specific location or individual's name related to governance or legal issues

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     bump
    -0.66
     splitting
    -0.66
     patches
    -0.65
     drinking
    -0.65
     splash
    -0.64
     Hack
    -0.63
     outputs
    -0.62
    heads
    -0.62
     split
    -0.62
    sid
    -0.61
    POSITIVE LOGITS
    arte
    5.06
    alde
    1.32
    enne
    1.14
    adena
    1.13
    arta
    1.12
    oche
    1.04
    artist
    1.03
    erva
    1.03
    atel
    1.01
    oso
    0.99
    Activations Density 0.030%

    No Known Activations