© 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. 10-RES-JB
    11. 1208
    Prev
    Next
    INDEX
    Explanations

    references to events or activities involving parties

    oai_token-act-pair · gpt-3.5-turbo

    references to parties or celebrations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Grizz
    -0.68
     Stafford
    -0.68
    abe
    -0.66
     Marketable
    -0.66
     Lif
    -0.65
     Pwr
    -0.64
    raham
    -0.64
    éĸ
    -0.63
    teenth
    -0.62
    obic
    -0.62
    POSITIVE LOGITS
    goers
    1.35
     planner
    0.92
    room
    0.89
     invitations
    0.88
     favors
    0.87
    ratulations
    0.81
     banter
    0.80
     dresses
    0.79
     planners
    0.77
     guests
    0.76
    Activations Density 0.030%

    No Known Activations