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. 14128
    Prev
    Next
    INDEX
    Explanations

    terms related to specific words, "Hour" and "Spectacular."

    oai_token-act-pair · gpt-3.5-turbo

    references to time-related concepts, particularly those associated with the term "Hour."

    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
     prov
    -0.76
    Iv
    -0.76
     capitals
    -0.71
     maize
    -0.71
     NK
    -0.70
     targ
    -0.70
     isEnabled
    -0.69
    zo
    -0.69
     graft
    -0.68
    iciary
    -0.67
    POSITIVE LOGITS
     Hour
    2.22
     Unlimited
    1.66
     Spect
    1.63
     Summers
    1.27
     eclipse
    1.23
     Und
    1.12
    frey
    1.05
     unborn
    1.05
     Unt
    1.05
    Hour
    1.02
    Activations Density 0.055%

    No Known Activations