© 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. 8-RES-JB
    11. 17795
    Prev
    Next
    INDEX
    Explanations

    phrases related to being filled or covered with something, such as messages, ads, or issues

    oai_token-act-pair · gpt-3.5-turbo

    descriptors and phrases indicating clutter or disorder

    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
    Not in Any Lists

    No Comments

    Negative Logits
    uber
    -0.75
     srfAttach
    -0.72
     compr
    -0.70
    IFF
    -0.69
    peer
    -0.67
    alt
    -0.67
    quickShipAvailable
    -0.67
    argument
    -0.66
    iggurat
    -0.65
    held
    -0.62
    POSITIVE LOGITS
     inbox
    0.78
    iously
    0.77
     coffers
    0.75
    icative
    0.72
     waters
    0.70
    uzz
    0.69
     dots
    0.67
    ious
    0.67
     Ples
    0.66
     with
    0.66
    Activations Density 0.164%

    No Known Activations