© 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. 10-RES-JB
    11. 2664
    Prev
    Next
    INDEX
    Explanations

    words/phrases ending with 'h'

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the letter "h" in the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Tyrann
    -0.66
     mammoth
    -0.66
     tremend
    -0.63
     tyrann
    -0.60
     terminal
    -0.59
     Tonight
    -0.59
     princ
    -0.58
     accountable
    -0.58
     congr
    -0.57
    ente
    -0.56
    POSITIVE LOGITS
    irts
    1.36
    adow
    1.18
    ooters
    1.16
    ooter
    1.08
    ield
    1.08
    IFT
    1.02
    oulder
    0.99
    ippers
    0.98
    owers
    0.98
    apers
    0.95
    Activations Density 0.030%

    No Known Activations