© 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. 2713
    Prev
    Next
    INDEX
    Explanations

    the word "just" followed by a number indicating the strength of connection

    oai_token-act-pair · gpt-3.5-turbo

    the word "just" in various contexts

    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
    lav
    -0.70
     plaintiff
    -0.69
    undai
    -0.65
     adversary
    -0.64
    antage
    -0.64
    leneck
    -0.64
    ixel
    -0.63
     contention
    -0.61
     foes
    -0.61
     adversaries
    -0.60
    POSITIVE LOGITS
    ifiable
    1.15
    ifications
    1.12
    ifi
    0.98
    if
    0.97
    ices
    0.90
    IFIC
    0.90
    IFIED
    0.83
     kidding
    0.83
    ifiers
    0.82
    itia
    0.82
    Activations Density 0.110%

    No Known Activations