© 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. 12390
    Prev
    Next
    INDEX
    Explanations

    phrases related to arguing a specific point or position

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "case" in various contexts

    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
    autical
    -0.81
    attery
    -0.81
    pez
    -0.75
    airs
    -0.74
    atomic
    -0.72
    elta
    -0.72
    orks
    -0.72
    assi
    -0.71
    CG
    -0.70
    onder
    -0.69
    POSITIVE LOGITS
    cases
    1.03
    case
    0.90
    backs
    0.79
     against
    0.73
    Against
    0.71
     Keen
    0.70
    back
    0.67
    BACK
    0.67
     Ruk
    0.66
    suit
    0.66
    Activations Density 0.033%

    No Known Activations