© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 1772
    Prev
    Next
    INDEX
    Explanations

    phrases related to abbreviations and terms related to specific fields or subjects

    oai_token-act-pair · gpt-3.5-turbo

    references to "AD" likely related to a specific topic or category, possibly indicating a focus on administrative details or announcements

    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
     htt
    -0.74
    andum
    -0.62
     externalToEVAOnly
    -0.61
    assetsadobe
    -0.60
    zzle
    -0.60
    drop
    -0.60
    otine
    -0.59
    href
    -0.58
     Fey
    -0.58
     Stronghold
    -0.57
    POSITIVE LOGITS
    VERTISEMENT
    1.25
    VERTIS
    1.18
    MIN
    1.11
    vantage
    0.98
    VIS
    0.92
    venture
    0.91
    ILY
    0.90
    iamond
    0.89
    DIS
    0.89
    DR
    0.89
    Activations Density 0.032%

    No Known Activations