© 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. 2-RES-JB
    11. 1299
    Prev
    Next
    INDEX
    Explanations

    references to the physical exercise "squatting"

    oai_token-act-pair · gpt-3.5-turbo

    references to squatting exercises and techniques

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Preferred
    -0.70
     Shore
    -0.64
     Strait
    -0.63
    ï¸ı
    -0.63
    displayText
    -0.62
    Adv
    -0.60
     Bond
    -0.59
     senses
    -0.59
     Learned
    -0.59
     HPV
    -0.58
    POSITIVE LOGITS
     squats
    1.06
     squat
    1.05
    stairs
    0.91
    lift
    0.81
    halla
    0.81
    chet
    0.78
    lifting
    0.75
    neck
    0.72
    alion
    0.72
    hest
    0.71
    Activations Density 0.011%

    No Known Activations