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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 3-RES-JB
    11. 11981
    Prev
    Next
    INDEX
    Explanations

    software and technology company names

    oai_token-act-pair · gpt-3.5-turbo

    specific technology company names, particularly those related to software and hardware industries

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ablishment
    -0.72
    utherford
    -0.70
    ulhu
    -0.70
    olkien
    -0.69
    Redditor
    -0.68
    avez
    -0.67
    eal
    -0.65
    usterity
    -0.64
    endants
    -0.64
    oppers
    -0.62
    POSITIVE LOGITS
     Corporation
    0.88
     Perkins
    0.78
    Mobil
    0.72
     Associates
    0.71
     Labs
    0.70
     executive
    0.69
    ites
    0.67
     Electronics
    0.66
     Sachs
    0.65
     Corp
    0.65
    Activations Density 0.013%

    No Known Activations