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

    Neuronpedia

    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. 7-RES-JB
    11. 2312
    Prev
    Next
    INDEX
    Explanations

    pronouns with verbs indicating action towards an object

    oai_token-act-pair · gpt-3.5-turbo

    references to an object or concept being acted upon or discussed repeatedly

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 7-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    7
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ãĥ©ãĥ³
    -0.65
    idth
    -0.62
     Frontier
    -0.62
     United
    -0.58
     Governments
    -0.58
     Mans
    -0.57
    mission
    -0.57
     Telecommunications
    -0.56
     Foreign
    -0.56
     Geneva
    -0.54
    POSITIVE LOGITS
    alian
    1.37
    self
    1.21
    unes
    1.21
    chy
    1.01
    iner
    0.90
    asca
    0.88
    ELF
    0.84
    geist
    0.83
    ueller
    0.80
    atic
    0.79
    Activations Density 0.168%

    No Known Activations