© 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. 21178
    Prev
    Next
    INDEX
    Explanations

    phrases related to qualities or characteristics, such as "strong and sensitive" or "flexible and lightweight."

    oai_token-act-pair · gpt-3.5-turbo

    co-occurrences and patterns in phrases that convey complexity or duality in various contexts

    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
    ļéĨĴ
    -0.73
     VIDEOS
    -0.73
    ESE
    -0.73
    ONSORED
    -0.72
    IRT
    -0.71
    atical
    -0.71
    ATS
    -0.71
    :(
    -0.70
    odcast
    -0.69
    éŃĶ
    -0.69
    POSITIVE LOGITS
     therefore
    1.04
     consequently
    0.94
     hence
    0.91
     thus
    0.90
    rogen
    0.89
     then
    0.88
    rogens
    0.84
     possibly
    0.81
     moreover
    0.76
     hopefully
    0.71
    Activations Density 0.716%

    No Known Activations