© 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. 3-RES-JB
    11. 541
    Prev
    Next
    INDEX
    Explanations

    adjectives describing a lack of awareness or experience

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "naive" and its variations in contexts questioning judgment and innocence

    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
    ŃĶ
    -0.80
    Downloadha
    -0.77
    alach
    -0.76
     srf
    -0.74
    interrupted
    -0.73
     contiguous
    -0.72
    hops
    -0.71
    ngth
    -0.71
    avez
    -0.70
    foreseen
    -0.69
    POSITIVE LOGITS
    te
    0.99
     naive
    0.98
    lings
    0.92
    sters
    0.86
    tle
    0.84
    ster
    0.82
    ted
    0.82
     innocence
    0.82
    ïve
    0.81
     naïve
    0.81
    Activations Density 0.013%

    No Known Activations