© 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. 3-RES-JB
    11. 9327
    Prev
    Next
    INDEX
    Explanations

    words related to regeneration and renewal

    oai_token-act-pair · gpt-3.5-turbo

    the mention of "reg" or related terms, indicating a focus on regulations or community guidelines

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    lihood
    -0.82
    WARE
    -0.77
    hower
    -0.76
    âĶģ
    -0.74
    Fine
    -0.73
     Fey
    -0.72
     Heck
    -0.71
     Trap
    -0.71
     Realms
    -0.69
    BOOK
    -0.68
    POSITIVE LOGITS
    rowth
    1.10
    inal
    1.08
    roup
    1.06
    gae
    1.05
    ressive
    1.04
    aining
    1.01
    arded
    0.97
    nant
    0.94
    ulated
    0.93
    ents
    0.93
    Activations Density 0.004%

    No Known Activations