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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 3-RES-JB
    6. 1971
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "construct" and related forms

    oai_token-act-pair · gpt-3.5-turbo

    various forms of the word "construct" and its derivatives

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Burk
    -0.71
    CVE
    -0.67
    eka
    -0.66
     beet
    -0.64
    EGA
    -0.63
     Moz
    -0.63
     Chili
    -0.61
    vana
    -0.61
    ILLE
    -0.61
    bid
    -0.60
    POSITIVE LOGITS
    ions
    1.28
    ivist
    1.26
    smanship
    1.19
    ively
    1.16
    ivism
    1.06
    Construct
    1.06
     construct
    1.01
    ivity
    0.95
     constructs
    0.95
    ors
    0.95
    Activations Density 0.019%

    No Known Activations