© 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. 16692
    Prev
    Next
    INDEX
    Explanations

    references to the name "Gard" or "garda", which could be related to a person's name or a specific entity

    oai_token-act-pair · gpt-3.5-turbo

    mentions of "Gard" and related terms

    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
    uyomi
    -0.82
    pson
    -0.80
    igslist
    -0.69
    izable
    -0.67
    phrine
    -0.64
    ãģį
    -0.63
    acca
    -0.61
    avez
    -0.61
    ignment
    -0.59
    BIP
    -0.59
    POSITIVE LOGITS
    llan
    1.01
    ener
    0.95
    inia
    0.95
    gard
    0.89
    folk
    0.86
    rell
    0.85
    leness
    0.85
    endale
    0.82
    ening
    0.82
    yne
    0.82
    Activations Density 0.038%

    No Known Activations