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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 9-RES-JB
    6. 2125
    Prev
    Next
    INDEX
    Explanations

    statements related to the concept of evidence and its role in defining reality

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 9-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    9
    Activation Function
    relu
    Embeds
    IFrame
    Link
    •Conspiracy theories
    'evidence and its role in defining reality'

    No Comments

    Negative Logits
     dubbed
    -0.83
     unexpectedly
    -0.81
     unusually
    -0.80
     billed
    -0.80
     previously
    -0.80
     scheduled
    -0.78
     somewhat
    -0.78
    abouts
    -0.77
     requested
    -0.75
     prompted
    -0.75
    POSITIVE LOGITS
     Therefore
    1.54
     Hence
    1.46
     Otherwise
    1.34
     Nobody
    1.33
     Neither
    1.32
     Likewise
    1.31
     Unless
    1.29
     Period
    1.29
     Everyone
    1.28
     Anyone
    1.28
    Activations Density 0.517%

    No Known Activations