© 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
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 3-RES-JB
    6. 15509
    Prev
    Next
    INDEX
    Explanations

    instances of words related to specific historical and scientific references, particularly focusing on names and terms related to war, medical conditions, and scientific discoveries

    oai_token-act-pair · gpt-3.5-turbo

    references to historical figures and events

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    arnaev
    -0.90
    intosh
    -0.89
    lass
    -0.84
    istani
    -0.84
    ointed
    -0.83
    abases
    -0.81
    rug
    -0.81
    nan
    -0.80
    uterte
    -0.79
    nets
    -0.79
    POSITIVE LOGITS
     gas
    0.85
     Claudia
    0.78
     helium
    0.78
     xen
    0.67
     tender
    0.67
     Gemini
    0.67
     atom
    0.66
     purity
    0.65
    hovah
    0.65
     casualty
    0.64
    Activations Density 0.025%

    No Known Activations