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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 0-RES-JB
    11. 23104
    Prev
    Next
    INDEX
    Explanations

    words related to statements of fact or attribution

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "are."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     itself
    -0.68
     herself
    -0.62
     himself
    -0.62
     dism
    -0.62
     QC
    -0.61
     myself
    -0.58
    EngineDebug
    -0.57
     rebounds
    -0.57
     serial
    -0.56
     manually
    -0.56
    POSITIVE LOGITS
    are
    4.24
    ARE
    2.34
    ares
    2.34
    aren
    1.94
    ared
    1.92
    were
    1.59
    aring
    1.59
    have
    1.42
    arer
    1.33
    aret
    1.31
    Activations Density 0.009%

    No Known Activations