© 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. 24163
    Prev
    Next
    INDEX
    Explanations

    phrases indicating tips, lessons, or learnings

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "less" and its variations, indicating a focus on comparisons or reductions

    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
    AE
    -0.67
     Unlock
    -0.65
    âĹ¼
    -0.63
    mberg
    -0.60
    Draft
    -0.60
     Reconstruction
    -0.59
    Move
    -0.59
     Culture
    -0.59
    å¯
    -0.59
    SU
    -0.59
    POSITIVE LOGITS
    ened
    1.03
    ening
    0.99
     than
    0.99
     fortunate
    0.92
    ons
    0.84
    ener
    0.78
    than
    0.77
    ens
    0.77
    ington
    0.73
     traveled
    0.71
    Activations Density 0.045%

    No Known Activations