Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 3-RES-JB
    6. 11739
    Prev
    Next
    INDEX
    Explanations

    phrases related to high temperature

    oai_token-act-pair · gpt-3.5-turbo

    the term "hot" in various contexts

    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
    ufact
    -0.88
    eca
    -0.77
    uther
    -0.75
    vous
    -0.75
    uthor
    -0.74
     Borders
    -0.68
    yss
    -0.68
    ARDIS
    -0.68
    ajor
    -0.67
    bernatorial
    -0.67
    POSITIVE LOGITS
     hot
    0.95
     hotter
    0.92
     hottest
    0.91
    hot
    0.89
     spots
    0.88
     stove
    0.86
    ened
    0.86
     sauce
    0.85
    ness
    0.84
    Hot
    0.81
    Activations Density 0.010%

    No Known Activations