© 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. 887
    Prev
    Next
    INDEX
    Explanations

    words related to fine powders or substances that can be used in various contexts

    oai_token-act-pair · gpt-3.5-turbo

    references to various types of powder

    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
    ĻĤ
    -0.88
    alian
    -0.79
    ocene
    -0.77
     Democr
    -0.74
    egal
    -0.71
    rian
    -0.71
    ovation
    -0.71
    IAN
    -0.67
    ŃĶ
    -0.67
    obal
    -0.64
    POSITIVE LOGITS
     Powder
    0.95
    powder
    0.94
    stals
    0.93
     powder
    0.87
     powd
    0.86
    flake
    0.83
    puff
    0.82
    roleum
    0.81
    bowl
    0.78
    bags
    0.76
    Activations Density 0.014%

    No Known Activations