© 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. 3-RES-JB
    11. 20928
    Prev
    Next
    INDEX
    Explanations

    references to soccer teams

    oai_token-act-pair · gpt-3.5-turbo

    references to football (soccer) teams

    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
    renheit
    -0.85
    untu
    -0.77
    enegger
    -0.72
    okia
    -0.70
    ksh
    -0.68
     MFT
    -0.67
    efined
    -0.67
    erous
    -0.65
    umerable
    -0.65
    opian
    -0.65
    POSITIVE LOGITS
    kick
    0.85
    club
    0.76
    por
    0.73
     acad
    0.73
     FC
    0.71
     midfielder
    0.71
     strugg
    0.70
     Jagu
    0.70
     ace
    0.70
    side
    0.69
    Activations Density 0.011%

    No Known Activations