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

    Neuronpedia

    Jacobian LensNEW
    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. 8-RES-JB
    11. 11223
    Prev
    Next
    INDEX
    Explanations

    words or phrases related to complex legal terms or actions

    oai_token-act-pair · gpt-3.5-turbo

    words related to complex concepts or relationships, especially those involving conflict or agreement

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    stones
    -0.64
    ħĭ
    -0.62
    Ctrl
    -0.61
    works
    -0.60
    AMI
    -0.60
    emis
    -0.59
    gdala
    -0.58
     Balls
    -0.58
    ington
    -0.58
    uesday
    -0.57
    POSITIVE LOGITS
    ably
    0.88
    ciating
    0.80
    ileged
    0.74
    ivable
    0.71
    ibly
    0.69
    isable
    0.69
    inately
    0.67
    icable
    0.67
    rencies
    0.66
    asion
    0.66
    Activations Density 0.118%

    No Known Activations