© 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. 11-RES-JB
    11. 5823
    Prev
    Next
    INDEX
    Explanations

    words related to actions or processes of various kinds, especially those involving movement or force

    oai_token-act-pair · gpt-3.5-turbo

    instances of the letters "tr" in various forms and contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     buckle
    -0.72
     hinge
    -0.67
     Mechdragon
    -0.66
     need
    -0.61
     rake
    -0.61
     minded
    -0.60
     rehe
    -0.60
     Nationwide
    -0.59
     spoiler
    -0.58
     calling
    -0.58
    POSITIVE LOGITS
    imester
    1.40
    unks
    1.36
    umps
    1.29
    icycle
    1.28
    udging
    1.27
    inity
    1.25
    ifling
    1.24
    acer
    1.22
    iceps
    1.21
    uces
    1.19
    Activations Density 0.010%

    No Known Activations