© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 4-RES-JB
    11. 2734
    Prev
    Next
    INDEX
    Explanations

    the concept of ownership or belonging

    oai_token-act-pair · gpt-3.5-turbo

    the concept of belonging

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

    No Comments

    Negative Logits
    olson
    -0.71
    imar
    -0.66
    airs
    -0.65
     helicop
    -0.63
     relay
    -0.62
    ork
    -0.61
    magic
    -0.61
     Sweeney
    -0.60
    ropolis
    -0.60
     turnaround
    -0.60
    POSITIVE LOGITS
     belong
    0.92
     belonged
    0.89
     belongs
    0.83
    itialized
    0.80
    ĪĴ
    0.79
     belonging
    0.77
    itors
    0.75
    BuyableInstoreAndOnline
    0.73
    alties
    0.72
    holder
    0.71
    Activations Density 0.010%

    No Known Activations