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

    Neuronpedia

    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. 1-RES-JB
    11. 13186
    Prev
    Next
    INDEX
    Explanations

    statements indicating reasons or justifications

    oai_token-act-pair · gpt-3.5-turbo

    multiple instances of the word "reasons."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     stocking
    -0.75
    ream
    -0.72
     puck
    -0.69
    esc
    -0.68
     Roller
    -0.67
    cycl
    -0.66
    yss
    -0.64
    ymph
    -0.61
    shr
    -0.59
     Pixie
    -0.59
    POSITIVE LOGITS
     why
    1.07
     reasons
    1.05
     WHY
    1.05
    why
    0.91
     justifying
    0.87
    resy
    0.80
     excuses
    0.79
    DragonMagazine
    0.79
     rationale
    0.78
     motivations
    0.78
    Activations Density 0.013%

    No Known Activations