© 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. 9056
    Prev
    Next
    INDEX
    Explanations

    references to specific video game titles including "Call of Duty."

    oai_token-act-pair · gpt-3.5-turbo

    references to the "Call of Duty" video game franchise

    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
     istg
    -0.73
     exha
    -0.72
     concentration
    -0.70
     ende
    -0.67
    cffff
    -0.66
     therape
    -0.65
     tremend
    -0.64
    ————
    -0.61
     vaginal
    -0.61
     centrif
    -0.60
    POSITIVE LOGITS
    backs
    1.31
    aghan
    1.11
    call
    1.03
     Call
    0.98
    igraph
    0.94
    Call
    0.94
    ers
    0.89
    aires
    0.87
    outs
    0.85
    oway
    0.84
    Activations Density 0.021%

    No Known Activations