© 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. 12881
    Prev
    Next
    INDEX
    Explanations

    instances where individuals or entities are being specifically trained for certain tasks or roles

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to training or instruction

    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
    FACE
    -0.81
    sm
    -0.72
    Minecraft
    -0.69
    eu
    -0.68
    esa
    -0.67
    fr
    -0.67
     Meier
    -0.66
    itus
    -0.66
    stuff
    -0.62
    issues
    -0.62
    POSITIVE LOGITS
     trained
    1.00
    anamo
    0.89
    anooga
    0.89
     veter
    0.87
    dinand
    0.84
     certific
    0.83
     professionals
    0.81
    DragonMagazine
    0.81
     instructors
    0.79
    arnaev
    0.79
    Activations Density 0.010%

    No Known Activations