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. 1-RES-JB
    6. 12402
    Prev
    Next
    INDEX
    Explanations

    mentions of countries or names that include the word 'ran'

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "ran" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ãĢIJ
    -0.81
    Merit
    -0.73
    ãĢ
    -0.68
    ðĿ
    -0.66
    º
    -0.66
    ĨĴ
    -0.65
     magnitude
    -0.64
    GMT
    -0.63
    é¾įå¥ij士
    -0.62
     aptly
    -0.61
    POSITIVE LOGITS
    ran
    1.23
    igans
    1.20
    vier
    1.13
    igan
    0.94
    quez
    0.90
    fo
    0.88
    stal
    0.85
    kered
    0.79
    mers
    0.77
    river
    0.76
    Activations Density 0.010%

    No Known Activations