© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 3-RES-JB
    11. 23994
    Prev
    Next
    INDEX
    Explanations

    information related to operating systems and computer hardware

    oai_token-act-pair · gpt-3.5-turbo

    references to different computer operating systems and hardware environments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    actionDate
    -0.92
    SHIP
    -0.83
    sbm
    -0.73
    forestation
    -0.72
    thumbnails
    -0.71
     mete
    -0.69
    rak
    -0.69
    ilings
    -0.67
    rahim
    -0.66
     Rasmussen
    -0.65
    POSITIVE LOGITS
     emulator
    0.79
     Vista
    0.78
     compatibility
    0.78
     OEM
    0.78
     Macintosh
    0.74
     browsers
    0.73
    Linux
    0.73
     DirectX
    0.72
     Linux
    0.72
     macOS
    0.72
    Activations Density 0.204%

    No Known Activations