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. 23753
    Prev
    Next
    INDEX
    Explanations

    phrases indicating an official statement or position being documented or recorded

    oai_token-act-pair · gpt-3.5-turbo

    phrases that indicate the concept of "record" and its various contexts

    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
    adin
    -0.89
     Franch
    -0.75
    hin
    -0.74
    issan
    -0.73
    tein
    -0.68
    bite
    -0.67
    tesy
    -0.67
    iege
    -0.66
     Lumpur
    -0.65
    creen
    -0.65
    POSITIVE LOGITS
     record
    1.04
    record
    0.99
     breaker
    0.91
     Record
    0.89
     records
    0.81
     holder
    0.80
     linkage
    0.78
     Holder
    0.78
     labels
    0.78
    breaking
    0.74
    Activations Density 0.024%

    No Known Activations