© 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 EvalsExportsinterp-engineNEWAPI 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. 16611
    Prev
    Next
    INDEX
    Explanations

    words related to items that are able to be acted upon or manipulated

    oai_token-act-pair · gpt-3.5-turbo

    words that describe capabilities or functions

    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
    krit
    -0.83
    erva
    -0.82
    arte
    -0.74
    akedown
    -0.70
    herry
    -0.69
    otaur
    -0.68
    itect
    -0.68
    aeper
    -0.68
    urer
    -0.67
    ebus
    -0.66
    POSITIVE LOGITS
    Redditor
    0.75
    nesses
    0.73
    pedia
    0.72
    Constructed
    0.69
    Characters
    0.67
    Ptr
    0.64
     Antiqu
    0.63
    lihood
    0.63
     lapse
    0.62
    nces
    0.61
    Activations Density 0.029%

    No Known Activations