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

    Neuronpedia

    APICircuit 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. 14263
    Prev
    Next
    INDEX
    Explanations

    cloud-related terms, especially the word "cloud"

    oai_token-act-pair · gpt-3.5-turbo

    references to cloud computing

    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
     hon
    -0.71
     Fas
    -0.67
     positional
    -0.65
     Golden
    -0.62
     recess
    -0.61
     Upper
    -0.60
     um
    -0.60
     Tommy
    -0.60
    omm
    -0.60
     meat
    -0.60
    POSITIVE LOGITS
    Cloud
    3.89
    cloud
    3.32
     Clouds
    1.26
     cloud
    1.24
     clouds
    1.22
     Cloud
    1.09
    Ocean
    1.01
    banks
    0.90
    BuyableInstoreAndOnline
    0.90
    Tube
    0.90
    Activations Density 0.020%

    No Known Activations