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. 9045
    Prev
    Next
    INDEX
    Explanations

    time-related words, especially those indicating movement forwards or backwards in time

    oai_token-act-pair · gpt-3.5-turbo

    references to transitions or movements, particularly of "forth" and "forward."

    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
     systems
    -0.60
     li
    -0.60
     Gig
    -0.58
     Hernandez
    -0.57
     ticks
    -0.57
     spect
    -0.56
     System
    -0.56
     Gy
    -0.56
     Seahawks
    -0.56
     system
    -0.56
    POSITIVE LOGITS
    forth
    5.45
    forward
    2.27
     Forth
    1.65
    theless
    1.25
    stairs
    1.11
    upon
    1.05
     forth
    1.02
    shall
    0.96
    soever
    0.95
    gow
    0.95
    Activations Density 0.016%

    No Known Activations