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. 0-RES-JB
    6. 19711
    Prev
    Next
    INDEX
    Explanations

    words related to defeat or failure

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "defeat" and its variations

    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
    XY
    -0.82
    sci
    -0.72
     Fuller
    -0.68
    hed
    -0.68
    ellen
    -0.66
     Signature
    -0.65
    hey
    -0.63
    eters
    -0.62
    ular
    -0.60
    reads
    -0.59
    POSITIVE LOGITS
     defeat
    3.86
     defeats
    2.78
     Defeat
    2.38
     victory
    2.14
     defeated
    2.03
     defeating
    2.01
     victories
    1.67
     triumph
    1.55
     vanquished
    1.53
     setback
    1.53
    Activations Density 0.019%

    No Known Activations