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

    Neuronpedia

    APISteerSAE EvalsBlog/PodcastNEWSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 5-RES-JB
    6. 16408
    Prev
    Next
    INDEX
    Explanations

    actions or tasks in the past tense

    oai_token-act-pair · gpt-3.5-turbo

    instances of the verbs "take" and "took," particularly in contexts involving duration or effort

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 5-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    5
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    luaj
    -0.87
    tor
    -0.72
    idding
    -0.71
    don
    -0.69
     Patron
    -0.68
    tions
    -0.66
    ramid
    -0.66
    lib
    -0.63
    mson
    -0.63
    present
    -0.62
    POSITIVE LOGITS
     aback
    0.82
    aways
    0.81
    ynthesis
    0.75
     breaths
    0.72
     steps
    0.72
     precedence
    0.72
     place
    0.69
     awhile
    0.68
     precautions
    0.67
    OTOS
    0.67
    Activations Density 0.033%

    No Known Activations