© 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. 20976
    Prev
    Next
    INDEX
    Explanations

    proper nouns, especially names

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the saxophone and related musical contexts

    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
    ousel
    -0.71
    leased
    -0.69
    quote
    -0.65
    assetsadobe
    -0.65
    izont
    -0.65
    inventoryQuantity
    -0.62
    SIGN
    -0.62
    ociate
    -0.62
     soluble
    -0.61
    20439
    -0.60
    POSITIVE LOGITS
    ophone
    1.20
     Sax
    1.07
    Sax
    1.06
    ophon
    1.04
    onian
    0.76
    ena
    0.73
    uner
    0.71
     Roh
    0.71
     sax
    0.71
    ovy
    0.71
    Activations Density 0.006%

    No Known Activations