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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 0-RES-JB
    6. 17732
    Prev
    Next
    INDEX
    Explanations

    phrases related to ice hockey and specifically the action of handling or playing with the puck

    oai_token-act-pair · gpt-3.5-turbo

    references to hockey pucks and defensemen

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Birmingham
    -0.75
     Exodus
    -0.74
    ophon
    -0.74
     Nile
    -0.73
     EE
    -0.72
     Davis
    -0.71
     Virgin
    -0.69
     Cotton
    -0.68
    amar
    -0.67
    ammy
    -0.66
    POSITIVE LOGITS
     puck
    3.12
     defenseman
    2.42
     Penguins
    2.32
     goaltender
    2.25
     linem
    2.25
     Capitals
    2.20
     goalt
    2.18
     Canucks
    2.10
     Sabres
    2.09
     Canadiens
    2.09
    Activations Density 0.039%

    No Known Activations