© 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. 2-RES-JB
    11. 5335
    Prev
    Next
    INDEX
    Explanations

    words related to the concept of impossibility

    oai_token-act-pair · gpt-3.5-turbo

    statements about the concept of impossibility

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    lov
    -0.77
    rief
    -0.76
    ript
    -0.74
    ta
    -0.74
    avia
    -0.71
    uana
    -0.69
    rika
    -0.68
    chen
    -0.68
    iang
    -0.67
     UCS
    -0.67
    POSITIVE LOGITS
     impossible
    0.90
     feas
    0.86
     adolesc
    0.86
    ossible
    0.84
     improbable
    0.84
     unnecess
    0.83
    terday
    0.81
     mosqu
    0.77
     suspic
    0.76
     conceivable
    0.75
    Activations Density 0.019%

    No Known Activations