© 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 EvalsExportsinterp-engineNEWAPI 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. 8-RES-JB
    11. 24192
    Prev
    Next
    INDEX
    Explanations

    complex structures or situations that are difficult to navigate or understand

    oai_token-act-pair · gpt-3.5-turbo

    metaphorical references to complexity and confusion in experiences or situations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    arers
    -0.84
    orers
    -0.82
    orthy
    -0.82
    thood
    -0.76
    acons
    -0.76
    arations
    -0.74
    ignty
    -0.72
    ouch
    -0.69
    akens
    -0.67
    ocard
    -0.66
    POSITIVE LOGITS
     maze
    1.25
     labyrinth
    1.01
     slog
    0.93
     convol
    0.92
     juggling
    0.91
     scramble
    0.91
     revolving
    0.89
     swirl
    0.89
     chore
    0.88
     cascade
    0.88
    Activations Density 0.341%

    No Known Activations