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

    Neuronpedia

    Jacobian LensNEW
    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. 4346
    Prev
    Next
    INDEX
    Explanations

    words related to physical impact or motion

    oai_token-act-pair · gpt-3.5-turbo

    verb forms associated with physical actions or sensations

    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
    alty
    -0.69
    ãĥīãĥ©
    -0.67
    chwitz
    -0.65
    uties
    -0.64
    OTOS
    -0.63
    rik
    -0.60
    izoph
    -0.60
    rador
    -0.60
    ilet
    -0.60
    Ľ
    -0.60
    POSITIVE LOGITS
    ishly
    0.93
    abouts
    0.80
    mouth
    0.77
     downhill
    0.74
    Downloadha
    0.70
     bloom
    0.69
    burst
    0.68
     havoc
    0.68
    interstitial
    0.66
    iness
    0.65
    Activations Density 0.139%

    No Known Activations