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. 8-RES-JB
    6. 7504
    Prev
    Next
    INDEX
    Explanations

    phrases related to machines or technical specifications

    oai_token-act-pair · gpt-3.5-turbo

    acronyms and abbreviations related to governance and compliance

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    •Prediction Features Layer 8 GPT2 Small
    •Examples of different Logit Weight Distributions - GPT2 Small Layer 8
    Prediction Feature Example
    •Happy Hour
    Prediction feature example

    No Comments

    Negative Logits
    hra
    -0.65
    ria
    -0.63
    hex
    -0.61
     Teg
    -0.61
    alys
    -0.60
    unia
    -0.59
    agen
    -0.59
    aga
    -0.59
    hu
    -0.59
    js
    -0.59
    POSITIVE LOGITS
    ING
    1.65
    ER
    1.56
    EST
    1.55
    URE
    1.51
    NING
    1.49
    LE
    1.48
    IES
    1.48
    ED
    1.48
    ERS
    1.47
    LIN
    1.47
    Activations Density 0.213%

    No Known Activations