© 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. 11-RES-JB
    11. 22064
    Prev
    Next
    INDEX
    Explanations

    references to specific individuals along with actions they are performing

    oai_token-act-pair · gpt-3.5-turbo

    pronouns and subjects indicating individuals or groups in a narrative context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    actionDate
    -0.92
    nown
    -0.84
    ERE
    -0.75
     Matter
    -0.73
    ĸļ
    -0.72
    ECTION
    -0.72
    ailability
    -0.69
     haven
    -0.67
    ASED
    -0.65
    etheless
    -0.65
    POSITIVE LOGITS
     prepares
    1.50
     learns
    1.49
     tries
    1.47
     executes
    1.45
     adjusts
    1.45
     performs
    1.44
     chooses
    1.44
     launches
    1.44
     decides
    1.44
     recovers
    1.42
    Activations Density 0.677%

    No Known Activations