© 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. 8-RES-JB
    11. 8231
    Prev
    Next
    INDEX
    Explanations

    the presence of the word "finish" or similar terms related to completing tasks

    oai_token-act-pair · gpt-3.5-turbo

    references to completing tasks or projects

    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
    uli
    -0.81
    elled
    -0.73
    hare
    -0.72
    resy
    -0.67
    andowski
    -0.65
    orum
    -0.65
    arsity
    -0.62
    uristic
    -0.62
    uru
    -0.61
    æ©Ł
    -0.60
    POSITIVE LOGITS
     touches
    0.84
     unfinished
    0.71
     composing
    0.69
     cleaning
    0.68
    strap
    0.68
    ners
    0.66
     mosqu
    0.65
     finishing
    0.64
     compiling
    0.63
     preparations
    0.63
    Activations Density 0.031%

    No Known Activations