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. 23586
    Prev
    Next
    INDEX
    Explanations

    terms related to competition or potential success in a specific context

    oai_token-act-pair · gpt-3.5-turbo

    references to a highly competitive or favored position in various contexts

    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
    rary
    -0.73
    ulk
    -0.67
    âĸĪ
    -0.63
     Breach
    -0.63
    hed
    -0.60
    âĸĪâĸĪ
    -0.60
    aves
    -0.59
     Pers
    -0.59
     Gest
    -0.59
     Glass
    -0.59
    POSITIVE LOGITS
     contender
    3.68
     contenders
    2.96
     frontrunner
    1.94
     challenger
    1.66
     competitor
    1.53
     challengers
    1.47
     winner
    1.31
     finalists
    1.29
     rival
    1.26
     candidate
    1.25
    Activations Density 0.014%

    No Known Activations