© 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. 2-RES-JB
    11. 11346
    Prev
    Next
    INDEX
    Explanations

    specific words related to different types of roles or characters, such as "Ranger," "Carpenter," "Slayer," "Ninja," and "Reaper."

    oai_token-act-pair · gpt-3.5-turbo

    terms related to characters and themes from fantasy and adventure narratives, particularly in gaming and film contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ilater
    -0.95
    itional
    -0.84
    itionally
    -0.84
    etimes
    -0.79
    mus
    -0.76
    rities
    -0.76
    elong
    -0.73
    vertisement
    -0.73
    iment
    -0.72
    isine
    -0.72
    POSITIVE LOGITS
     Runner
    0.78
     Slayer
    0.77
     Scout
    0.75
     Robo
    0.74
     Spy
    0.74
    llan
    0.72
     Worker
    0.71
    Gear
    0.71
    ãĤº
    0.71
     Drone
    0.69
    Activations Density 0.028%

    No Known Activations