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. 8-RES-JB
    6. 15210
    Prev
    Next
    INDEX
    Explanations

    instances of the phrase "used to"

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "used" in various contexts, particularly relating to past habits or conditions

    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
     achie
    -0.66
     Ascend
    -0.66
     undertaking
    -0.63
    antle
    -0.62
    olulu
    -0.61
    ignt
    -0.60
    merce
    -0.60
     awarding
    -0.58
    nesty
    -0.57
    vance
    -0.57
    POSITIVE LOGITS
    fully
    1.05
    FUL
    0.91
     condoms
    0.87
    fulness
    0.84
     sparing
    0.83
    ãĥ¼ãĥĨãĤ£
    0.83
     tyres
    0.72
    ful
    0.70
    full
    0.68
     tires
    0.67
    Activations Density 0.074%

    No Known Activations