© 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. 5493
    Prev
    Next
    INDEX
    Explanations

    the modal verb "would" followed by a verb, potentially indicative of future actions or hypothetical situations

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "would."

    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
    holder
    -0.68
    rylic
    -0.67
    values
    -0.65
    meta
    -0.65
    resources
    -0.63
    isted
    -0.61
    regulated
    -0.61
    basketball
    -0.60
     Modified
    -0.60
    engineering
    -0.60
    POSITIVE LOGITS
     be
    1.13
    ĸļ
    1.11
     surely
    1.00
     inevitably
    0.99
     doubtless
    0.98
     never
    0.95
     undoubtedly
    0.93
     eventually
    0.92
     suffice
    0.89
     probably
    0.89
    Activations Density 0.156%

    No Known Activations