© 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. 1-RES-JB
    11. 16154
    Prev
    Next
    INDEX
    Explanations

    words related to legal agreements or treaties

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the word "words" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    EV
    -0.76
    LV
    -0.76
     stage
    -0.72
    Previous
    -0.68
    CVE
    -0.67
    BUS
    -0.66
    GY
    -0.65
    ASE
    -0.63
    Ku
    -0.63
     phase
    -0.63
    POSITIVE LOGITS
    ords
    1.54
    poons
    0.97
    hip
    0.93
    quartered
    0.89
    hower
    0.88
    omething
    0.88
    uggest
    0.80
    poon
    0.79
    orts
    0.76
    ĸļ
    0.75
    Activations Density 0.004%

    No Known Activations