© 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. 1170
    Prev
    Next
    INDEX
    Explanations

    words related to an abbreviation "Op" followed by a single character or a number, with activations of 9 or 10

    oai_token-act-pair · gpt-3.5-turbo

    references to a specific type of medication or pharmaceutical agent

    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
    ��
    -0.70
    ãĤ¨ãĥ«
    -0.69
    BILITY
    -0.68
    ा
    -0.66
     waive
    -0.64
    asher
    -0.64
    ictional
    -0.63
    usters
    -0.63
     guarant
    -0.62
     rewarded
    -0.61
    POSITIVE LOGITS
     Op
    3.94
    Op
    2.73
     op
    2.14
     OP
    1.50
    op
    1.31
     Open
    1.22
     Ops
    1.20
     Oper
    1.15
    OP
    1.11
     opium
    1.10
    Activations Density 0.009%

    No Known Activations