© 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. 7-RES-JB
    11. 17884
    Prev
    Next
    INDEX
    Explanations

    references to media consumption, streaming, and device functionalities

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to streaming media and digital content consumption

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 7-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    7
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    stood
    -0.80
    peror
    -0.69
     Leh
    -0.68
    judicial
    -0.67
     Dak
    -0.66
     condem
    -0.65
    agall
    -0.64
    lethal
    -0.64
     tactics
    -0.64
     ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³
    -0.64
    POSITIVE LOGITS
    netflix
    0.94
     directories
    0.83
    Netflix
    0.81
     playlist
    0.80
    ongs
    0.80
     Gmail
    0.80
     Spotify
    0.80
     metadata
    0.79
    metadata
    0.79
     subreddits
    0.78
    Activations Density 0.522%

    No Known Activations