© 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. 3-RES-JB
    11. 23508
    Prev
    Next
    INDEX
    Explanations

    statistical and medical terms related to research and analysis

    oai_token-act-pair · gpt-3.5-turbo

    data related to medical statistics and abortion rates

    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
     rooft
    -0.78
    oil
    -0.71
    Shut
    -0.71
    çīĪ
    -0.69
    boats
    -0.68
     rust
    -0.68
    NEWS
    -0.67
     Armageddon
    -0.67
    atche
    -0.67
    Priv
    -0.67
    POSITIVE LOGITS
     (âĪĴ
    1.13
     statistically
    1.12
     covari
    1.06
     quantitative
    1.04
     âĪ¼
    1.02
     depressive
    1.01
     endogenous
    1.00
     clinically
    1.00
     maternal
    0.99
     âī¥
    0.98
    Activations Density 0.973%

    No Known Activations