© 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. 16534
    Prev
    Next
    INDEX
    Explanations

    references to specific names, potentially related to sports figures or individuals involved in legal proceedings

    oai_token-act-pair · gpt-3.5-turbo

    references to specific basketball players and their performance

    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
    stract
    -0.75
     Lumpur
    -0.74
    rals
    -0.69
     conflicts
    -0.68
    abwe
    -0.66
     Mechanical
    -0.65
    ktop
    -0.65
     Stur
    -0.63
    ropolitan
    -0.63
     Warfare
    -0.63
    POSITIVE LOGITS
    glass
    0.98
    sonian
    0.87
    inkle
    0.79
    spir
    0.79
    lihood
    0.78
    --------------------------------
    0.73
    artz
    0.71
    è¯
    0.70
    ibles
    0.69
    encers
    0.68
    Activations Density 0.021%

    No Known Activations