Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 21896
    Prev
    Next
    INDEX
    Explanations

    references to occurrences happening repeatedly or at short intervals

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "frequent."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Frankenstein
    -0.66
    igsaw
    -0.66
    here
    -0.65
    hood
    -0.63
    frog
    -0.61
    oslov
    -0.59
    kamp
    -0.58
    ettle
    -0.58
    enstein
    -0.58
     Reborn
    -0.57
    POSITIVE LOGITS
     frequent
    3.72
    requent
    2.15
     occasional
    1.77
     frequently
    1.70
     continual
    1.50
     periodic
    1.47
     regularly
    1.46
     frequ
    1.46
     regular
    1.45
     recurring
    1.44
    Activations Density 0.008%

    No Known Activations