© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 4-RES-JB
    11. 702
    Prev
    Next
    INDEX
    Explanations

    terms related to processes, actions, and decision-making

    oai_token-act-pair · gpt-3.5-turbo

    phrases and concepts related to decision-making and consequences

    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
    acion
    -0.59
    chal
    -0.54
    venge
    -0.52
    mast
    -0.52
     earthqu
    -0.51
     JJ
    -0.49
    amic
    -0.48
    anus
    -0.47
    reet
    -0.47
    angel
    -0.47
    POSITIVE LOGITS
     proponents
    0.66
     policymakers
    0.62
     advocates
    0.61
     nonetheless
    0.59
     inherently
    0.59
     typically
    0.58
     anecd
    0.58
     shifting
    0.57
     researchers
    0.56
    typically
    0.56
    Activations Density 1.629%

    No Known Activations