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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 4-RES-JB
    6. 3617
    Prev
    Next
    INDEX
    Explanations

    moments or instances within a text

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "moments" 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
    achev
    -0.82
    ply
    -0.76
    sonian
    -0.76
    raid
    -0.74
    GE
    -0.73
    lli
    -0.71
    CAST
    -0.70
    vacc
    -0.70
    XY
    -0.70
    sheet
    -0.70
    POSITIVE LOGITS
     moments
    1.23
     Moments
    1.03
     glimps
    0.95
    hift
    0.92
    aneously
    0.88
     moment
    0.86
     milestones
    0.83
     reminders
    0.82
     bouts
    0.78
    aneous
    0.78
    Activations Density 0.018%

    No Known Activations