© 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 EvalsExportsinterp-engineNEWAPI 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. 18515
    Prev
    Next
    INDEX
    Explanations

    phrases related to catastrophic events or outcomes

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "devastating" and its variations related to impactful negative events

    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
    cius
    -0.98
    ploma
    -0.76
    rity
    -0.68
    cling
    -0.66
    uties
    -0.66
    annis
    -0.66
    ublic
    -0.65
    ioch
    -0.64
    porter
    -0.64
    utation
    -0.64
    POSITIVE LOGITS
     blow
    0.95
     devast
    0.94
    ly
    0.93
     devastation
    0.90
     havoc
    0.87
     blows
    0.86
    ãĥ¼ãĥĨ
    0.82
    oire
    0.81
     shocks
    0.79
     impacts
    0.77
    Activations Density 0.013%

    No Known Activations