© 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. 4-RES-JB
    11. 21972
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal convictions

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "conviction" and its related forms

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

    No Comments

    Negative Logits
    orns
    -0.75
    ÃŁ
    -0.74
    hod
    -0.73
    hens
    -0.72
    esta
    -0.72
    gm
    -0.71
    kef
    -0.70
    hy
    -0.70
    oss
    -0.69
     neighbours
    -0.68
    POSITIVE LOGITS
     convictions
    1.05
     conviction
    1.01
    eering
    0.86
     convict
    0.81
    icts
    0.76
    iction
    0.74
    alties
    0.74
     convicted
    0.73
     Bonds
    0.73
     Coil
    0.73
    Activations Density 0.012%

    No Known Activations