Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 22135
    Prev
    Next
    INDEX
    Explanations

    phrases related to beginning or starting events or actions

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "opening" in various contexts

    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
    ahu
    -0.83
    soever
    -0.75
    STON
    -0.73
     Died
    -0.70
    conom
    -0.68
    vere
    -0.66
    abouts
    -0.65
    xual
    -0.64
    raq
    -0.63
     fared
    -0.63
    POSITIVE LOGITS
     opening
    3.67
     Opening
    2.74
     openings
    2.47
    Opening
    2.42
    opening
    2.32
     closing
    2.28
     opener
    1.92
     opened
    1.73
     opens
    1.64
     Closing
    1.53
    Activations Density 0.021%

    No Known Activations