© 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
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 11-RES-JB
    6. 7813
    Prev
    Next
    INDEX
    Explanations

    verbs related to beginning or initiating action

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to beginning or starting anew

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    obi
    -0.73
    ographics
    -0.67
    Origin
    -0.67
    pedia
    -0.61
    pha
    -0.60
    imon
    -0.59
    alth
    -0.58
    ashi
    -0.57
    icator
    -0.56
    icons
    -0.55
    POSITIVE LOGITS
     anew
    1.29
     af
    0.94
     raining
    0.93
     behaving
    0.92
     worrying
    0.89
     experimenting
    0.87
     repairing
    0.84
     thinking
    0.84
     preparing
    0.84
     messing
    0.83
    Activations Density 0.069%

    No Known Activations