© 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. 0-RES-JB
    11. 5028
    Prev
    Next
    INDEX
    Explanations

    mentions of radioactive materials

    oai_token-act-pair · gpt-3.5-turbo

    references to radioactive materials

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Pavilion
    -0.80
    BOOK
    -0.79
    accompan
    -0.76
    aura
    -0.76
     compan
    -0.76
    WS
    -0.72
     Refuge
    -0.69
     Gustav
    -0.69
     Tactics
    -0.69
    INESS
    -0.66
    POSITIVE LOGITS
     radioactive
    1.66
     Rot
    1.47
     pri
    1.28
     Leicester
    1.23
     Kling
    1.23
     Hubble
    1.19
     Klingon
    1.16
    ajor
    1.15
     Slov
    1.12
    Rot
    1.09
    Activations Density 0.052%

    No Known Activations