© 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. 23112
    Prev
    Next
    INDEX
    Explanations

    usage of the word "drop."

    oai_token-act-pair · gpt-3.5-turbo

    references to declines or decreases in quantity or value

    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
     Pamela
    -0.74
    uthor
    -0.70
     Franch
    -0.69
    amia
    -0.68
    PAC
    -0.66
    powers
    -0.66
    olitan
    -0.65
    verning
    -0.64
    central
    -0.63
    Understanding
    -0.62
    POSITIVE LOGITS
     drop
    3.69
     drops
    2.54
     Drop
    2.53
    drop
    2.46
     dropping
    2.12
    Drop
    2.12
     dropped
    2.02
     Drops
    1.95
    drops
    1.60
     plunge
    1.49
    Activations Density 0.017%

    No Known Activations