© 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. 2435
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a change or event that occurred before a certain point in time

    oai_token-act-pair · gpt-3.5-turbo

    the word "Before" and its various contexts in relation to events

    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
     perpet
    -0.76
     fault
    -0.67
     transl
    -0.63
     fest
    -0.61
     maxim
    -0.60
     minim
    -0.60
     trivial
    -0.59
     pus
    -0.58
     buck
    -0.58
     cent
    -0.57
    POSITIVE LOGITS
     Before
    3.11
    Before
    2.02
    before
    1.80
     BEFORE
    1.76
     Previously
    1.73
     Prior
    1.67
     Until
    1.64
     During
    1.59
     Without
    1.58
     After
    1.56
    Activations Density 0.015%

    No Known Activations