© 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. 0-RES-JB
    11. 19245
    Prev
    Next
    INDEX
    Explanations

    verbs related to tasks or actions of completing or finishing something

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to the act of completing tasks or achieving goals

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ounce
    -0.68
    hem
    -0.66
    ened
    -0.66
    alled
    -0.65
     heard
    -0.64
     opened
    -0.63
     burst
    -0.60
    ream
    -0.60
    ise
    -0.60
     stand
    -0.59
    POSITIVE LOGITS
     completing
    2.34
     achieving
    2.14
     maintaining
    2.08
     obtaining
    2.07
     constructing
    2.05
     navigating
    1.94
     discovering
    1.93
     implementing
    1.93
     acquiring
    1.93
     designing
    1.92
    Activations Density 0.067%

    No Known Activations