© 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. 3-RES-JB
    11. 24160
    Prev
    Next
    INDEX
    Explanations

    phrases related to morale, cohesion, and organizational behavior

    oai_token-act-pair · gpt-3.5-turbo

    terms related to morale and group cohesion

    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
    cart
    -0.72
    Interstitial
    -0.72
    notes
    -0.71
    Cart
    -0.69
    apor
    -0.69
    ault
    -0.69
    Woman
    -0.68
    forward
    -0.68
    aways
    -0.68
    Redditor
    -0.67
    POSITIVE LOGITS
     morale
    1.27
     guiActiveUn
    0.85
    utenant
    0.85
    izational
    0.80
     veter
    0.79
    phis
    0.78
     cohesion
    0.77
    ilitary
    0.76
    seless
    0.75
     reluct
    0.74
    Activations Density 0.030%

    No Known Activations