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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 4-RES-JB
    6. 21768
    Prev
    Next
    INDEX
    Explanations

    terms and concepts related to neural networks, including keywords like "neural networks," "visual recognition," and "neural arrays."

    oai_token-act-pair · gpt-3.5-turbo

    topics related to neural networks and their applications

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

    No Comments

    Negative Logits
     Kuwait
    -0.77
    ::::::::
    -0.76
     FSA
    -0.73
     amen
    -0.66
    ][/
    -0.64
    lv
    -0.64
     Khe
    -0.64
     Kens
    -0.63
    ufact
    -0.63
     unmarried
    -0.63
    POSITIVE LOGITS
     algorithms
    1.27
     theorem
    1.24
     algorithm
    1.15
    ministic
    1.12
     equations
    1.03
     computation
    1.01
     computational
    0.99
     Turing
    0.99
     comput
    0.98
     graphene
    0.98
    Activations Density 0.232%

    No Known Activations