© 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. 2-RES-JB
    11. 22583
    Prev
    Next
    INDEX
    Explanations

    phrases related to causing harm or destruction

    oai_token-act-pair · gpt-3.5-turbo

    references to acts of killing

    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
     Collider
    -0.76
    rial
    -0.73
    BuyableInstoreAndOnline
    -0.73
    umn
    -0.67
    Grad
    -0.66
     concess
    -0.65
    arity
    -0.65
    anwhile
    -0.64
    provided
    -0.62
     Celest
    -0.61
    POSITIVE LOGITS
     spree
    0.89
    icides
    0.88
    houses
    0.82
    killer
    0.81
     kill
    0.79
     killing
    0.79
    joy
    0.78
    switch
    0.78
    icide
    0.76
     killers
    0.75
    Activations Density 0.025%

    No Known Activations