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. 8-RES-JB
    6. 15397
    Prev
    Next
    INDEX
    Explanations

    phrases related to pointing out specific issues or concerns in various contexts

    oai_token-act-pair · gpt-3.5-turbo

    phrases that indicate a focus or emphasis on the word "here," suggesting a central concept or point of discussion

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     parap
    -0.66
    catentry
    -0.59
    visors
    -0.58
    emet
    -0.57
     corpses
    -0.55
    uously
    -0.54
    ams
    -0.54
     Gujar
    -0.53
     Bund
    -0.53
     warranties
    -0.53
    POSITIVE LOGITS
    abouts
    1.61
    tical
    1.54
    tics
    1.51
    tic
    1.21
    with
    0.82
     guiActiveUn
    0.78
    upon
    0.77
    after
    0.72
    ody
    0.71
    about
    0.71
    Activations Density 0.060%

    No Known Activations