© 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. 2515
    Prev
    Next
    INDEX
    Explanations

    dates written in the format of year followed by two digits as related to historical events

    oai_token-act-pair · gpt-3.5-turbo

    specific patterns or sequences of numbers, particularly the repeated occurrence of '19' followed by various digits

    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
     voic
    -0.80
     minions
    -0.68
     Pwr
    -0.64
     voice
    -0.64
    kson
    -0.64
     silence
    -0.64
     tech
    -0.64
     beat
    -0.63
     ranc
    -0.63
     pudding
    -0.63
    POSITIVE LOGITS
    19
    3.28
    18
    2.30
    1959
    2.22
    1945
    2.13
    1920
    2.11
    1950
    2.07
    17
    2.04
    1963
    1.96
    1960
    1.92
    21
    1.91
    Activations Density 0.011%

    No Known Activations