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

    Neuronpedia

    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. 8-RES-JB
    11. 7800
    Prev
    Next
    INDEX
    Explanations

    the word "Espionage" in various contexts

    oai_token-act-pair · gpt-3.5-turbo

    references to "Esp" or variations of it, potentially indicating themes related to espionage or entities with similar nomenclature

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    folk
    -0.84
    ĸļ
    -0.80
    İĭ
    -0.76
    umeric
    -0.76
    sbm
    -0.74
    gling
    -0.72
    EStreamFrame
    -0.68
    Redditor
    -0.66
    creen
    -0.66
    IFIED
    -0.65
    POSITIVE LOGITS
    resso
    1.29
    ionage
    1.22
    encer
    1.06
    acial
    0.97
    irit
    0.93
    ending
    0.88
    inda
    0.84
    ences
    0.84
    encers
    0.82
    acio
    0.81
    Activations Density 0.048%

    No Known Activations