© 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. 4-RES-JB
    11. 22624
    Prev
    Next
    INDEX
    Explanations

    references to TV shows, particularly "Breaking Bad" and "The Walking Dead"

    oai_token-act-pair · gpt-3.5-turbo

    titles or names related to popular TV shows, particularly those with "Walking" and "Breaking" in the title

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

    No Comments

    Negative Logits
    ional
    -0.87
    haps
    -0.84
    CLUD
    -0.76
    ited
    -0.75
    iting
    -0.74
    rss
    -0.73
    ificial
    -0.73
    nings
    -0.72
    rar
    -0.72
    lled
    -0.72
    POSITIVE LOGITS
     Walking
    1.06
     Dead
    1.01
     Redemption
    0.98
     Companion
    0.98
     Compan
    0.93
     Wire
    0.92
     Actress
    0.85
     Expend
    0.85
     Agent
    0.85
     Masquerade
    0.84
    Activations Density 0.031%

    No Known Activations