© 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. 0-RES-JB
    11. 2715
    Prev
    Next
    INDEX
    Explanations

    titles or phrases including the prefix "Tw" followed by a number

    oai_token-act-pair · gpt-3.5-turbo

    references to "Twilight" or related components

    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
     Berman
    -0.76
     Rico
    -0.67
     garbage
    -0.65
     NEXT
    -0.65
     SAR
    -0.65
     cleaning
    -0.64
     membrane
    -0.64
     DOI
    -0.64
     depreciation
    -0.63
     scaven
    -0.62
    POSITIVE LOGITS
    Tw
    3.74
    tw
    2.49
    TW
    1.62
     Twisted
    1.53
     Twitch
    1.46
    twitch
    1.45
     Twilight
    1.44
     Twin
    1.44
     Tw
    1.34
     Twins
    1.22
    Activations Density 0.024%

    No Known Activations