© 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. 20177
    Prev
    Next
    INDEX
    Explanations

    explanations or reasons for situations or behaviors

    oai_token-act-pair · gpt-3.5-turbo

    phrases or questions seeking to explain reasons or justifications

    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
     Roller
    -0.79
    rop
    -0.69
    Zone
    -0.68
    shaw
    -0.67
    ymph
    -0.66
    bows
    -0.65
    amps
    -0.65
    lator
    -0.63
     Riders
    -0.63
    trop
    -0.62
    POSITIVE LOGITS
    soever
    1.08
     why
    1.00
     WHY
    0.91
    why
    0.90
    iterranean
    0.78
     exactly
    0.74
    ricanes
    0.73
    Why
    0.73
    ihad
    0.71
    icago
    0.69
    Activations Density 0.037%

    No Known Activations