© 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. 5970
    Prev
    Next
    INDEX
    Explanations

    references to physical fitness and exercise, specifically focusing on squatting and related activities

    oai_token-act-pair · gpt-3.5-turbo

    statements regarding physical training and exercise recommendations

    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
    Bloom
    -0.95
     Sparrow
    -0.81
     Dahl
    -0.79
    pei
    -0.77
    onymous
    -0.76
     veto
    -0.75
    Property
    -0.73
    Pand
    -0.73
     incest
    -0.73
     cens
    -0.73
    POSITIVE LOGITS
     workouts
    1.87
     workout
    1.83
     aerobic
    1.72
     athletes
    1.67
     cardio
    1.63
     gym
    1.63
     Athlet
    1.58
     fitness
    1.57
     creatine
    1.57
     Fitness
    1.54
    Activations Density 0.555%

    No Known Activations