© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 8696
    Prev
    Next
    INDEX
    Explanations

    terms related to objects or tools that are practical or useful

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "handy" and its context within the document

    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
    inction
    -0.72
    ophy
    -0.70
    eree
    -0.70
    verson
    -0.70
    orous
    -0.69
    YL
    -0.68
    onde
    -0.66
    eming
    -0.65
    Rush
    -0.65
    192
    -0.65
    POSITIVE LOGITS
     handy
    1.07
    ãĥ¼ãĥĨ
    0.90
    ãĥķãĤ¡
    0.79
     drawer
    0.79
     tricks
    0.78
     nifty
    0.75
    glers
    0.72
    hooting
    0.70
    tips
    0.70
     referen
    0.69
    Activations Density 0.005%

    No Known Activations