Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 24327
    Prev
    Next
    INDEX
    Explanations

    terms related to medical conditions, specifically anemia

    oai_token-act-pair · gpt-3.5-turbo

    medical conditions related to blood disorders

    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
    going
    -0.76
    boat
    -0.70
    Neigh
    -0.68
     Zig
    -0.67
    EStreamFrame
    -0.67
    nick
    -0.67
    Way
    -0.66
    Adams
    -0.66
     Woody
    -0.66
    alty
    -0.66
    POSITIVE LOGITS
    emia
    3.79
    uria
    1.60
     despair
    1.16
    ei
    1.12
    esthesia
    0.89
    uo
    0.89
    uel
    0.85
     coma
    0.84
    enza
    0.84
     Hebdo
    0.82
    Activations Density 0.038%

    No Known Activations