© 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. 4-RES-JB
    11. 3669
    Prev
    Next
    INDEX
    Explanations

    measurements related to size and dimensions, specifically focusing on measurements indicated in millimeters

    oai_token-act-pair · gpt-3.5-turbo

    measurements specified in millimeters

    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
    comes
    -0.81
    lez
    -0.77
    come
    -0.77
    hran
    -0.75
    peak
    -0.73
     toile
    -0.69
    matched
    -0.68
    eve
    -0.68
    GROUND
    -0.65
     PubMed
    -0.64
    POSITIVE LOGITS
     diameter
    0.98
    ORPG
    0.90
    mmm
    0.87
    ueller
    0.87
    ooth
    0.86
    itted
    0.80
    uzzle
    0.79
    endment
    0.76
    oves
    0.74
    oses
    0.74
    Activations Density 0.011%

    No Known Activations