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. 4-RES-JB
    6. 5372
    Prev
    Next
    INDEX
    Explanations

    sentences that express significant or impactful statements

    oai_token-act-pair · gpt-3.5-turbo

    sentences that end with a period

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

    No Comments

    Negative Logits
    onde
    -0.89
     compr
    -0.79
    thal
    -0.78
    onga
    -0.78
    ascus
    -0.77
    anus
    -0.77
    culus
    -0.75
    ogle
    -0.73
    emaker
    -0.73
    zel
    -0.73
    POSITIVE LOGITS
     Firstly
    1.26
     Especially
    1.11
     Whether
    1.11
     Unless
    1.08
     Though
    1.08
     Although
    1.06
     Perhaps
    1.06
     Whereas
    1.06
     While
    1.04
     Suppose
    1.02
    Activations Density 0.955%

    No Known Activations