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. 5704
    Prev
    Next
    INDEX
    Explanations

    references to a specific entity or concept, specifically related to real estate

    oai_token-act-pair · gpt-3.5-turbo

    references to the football club Real Madrid

    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
    bowl
    -0.84
    è¦ļéĨĴ
    -0.74
    abus
    -0.65
    isphere
    -0.65
    ribe
    -0.65
    hower
    -0.64
     buggy
    -0.64
    auld
    -0.63
    owl
    -0.63
     funnel
    -0.62
    POSITIVE LOGITS
     Real
    3.67
    Real
    2.99
     REAL
    2.21
    real
    1.95
     real
    1.75
     Reality
    1.47
    reality
    1.29
     realism
    1.29
     Sporting
    1.26
     reality
    1.15
    Activations Density 0.021%

    No Known Activations