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. 3589
    Prev
    Next
    INDEX
    Explanations

    terms related to things that require a high level of effort, focus, or attention to detail

    oai_token-act-pair · gpt-3.5-turbo

    references to high levels of intensity in various contexts

    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
    rule
    -0.73
    bows
    -0.69
    aurus
    -0.69
    dale
    -0.68
     buckle
    -0.67
     Kraken
    -0.66
    cy
    -0.64
    lad
    -0.64
    li
    -0.63
    ness
    -0.63
    POSITIVE LOGITS
     intensive
    3.77
    intensive
    2.37
     intense
    1.45
     stren
    1.22
     rigorous
    1.21
     extensive
    1.12
     intensely
    1.11
     intens
    1.10
     intensified
    1.08
     ardu
    1.07
    Activations Density 0.014%

    No Known Activations