© 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. 8-RES-JB
    11. 17981
    Prev
    Next
    INDEX
    Explanations

    phrases related to political allies

    oai_token-act-pair · gpt-3.5-turbo

    references to allies and their relationships or status in various contexts

    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
    ifax
    -0.75
    SpaceEngineers
    -0.74
    acco
    -0.74
    fitting
    -0.71
    plet
    -0.71
    nit
    -0.71
    Interstitial
    -0.69
    teen
    -0.69
    erick
    -0.66
    iday
    -0.65
    POSITIVE LOGITS
     allies
    0.81
     milit
    0.80
     ally
    0.78
     Allies
    0.78
     Ally
    0.77
     flank
    0.72
     Mobil
    0.70
     allied
    0.70
     foe
    0.69
    hips
    0.68
    Activations Density 0.044%

    No Known Activations