© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 1855
    Prev
    Next
    INDEX
    Explanations

    references to the word "Google" within contexts

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the company Google

    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
     Tall
    -0.74
     continuity
    -0.72
     Lar
    -0.71
     II
    -0.71
     sy
    -0.71
     Division
    -0.70
     chamber
    -0.67
     pregnant
    -0.67
     bang
    -0.67
     range
    -0.66
    POSITIVE LOGITS
    Google
    3.55
    google
    1.94
     Google
    1.77
    Microsoft
    1.77
     google
    1.69
    Reddit
    1.63
    Amazon
    1.61
    Android
    1.61
    Samsung
    1.58
    LinkedIn
    1.57
    Activations Density 0.013%

    No Known Activations