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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlog/PodcastSlackPrivacy & 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. 15727
    Prev
    Next
    INDEX
    Explanations

    instances of the term "record" followed by either a noun or an adjective

    oai_token-act-pair · gpt-3.5-turbo

    phrases referring to record levels or highs in various contexts

    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
    adin
    -0.87
    hiro
    -0.81
    issan
    -0.76
    wich
    -0.73
    cius
    -0.69
    iant
    -0.69
    apa
    -0.69
     Franch
    -0.68
    iege
    -0.68
    pheus
    -0.68
    POSITIVE LOGITS
     record
    1.00
    record
    0.95
     breaker
    0.90
     Record
    0.87
    keeping
    0.86
    breaking
    0.84
     records
    0.81
     holder
    0.79
    downs
    0.77
    recorded
    0.76
    Activations Density 0.028%

    No Known Activations