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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 4-RES-JB
    6. 518
    Prev
    Next
    INDEX
    Explanations

    references to specific sources or documents

    oai_token-act-pair · gpt-3.5-turbo

    closing brackets and punctuation marks in lists or references

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

    No Comments

    Negative Logits
    »Ĵ
    -0.69
     sag
    -0.65
    minster
    -0.63
    ppy
    -0.63
    pping
    -0.61
    chuk
    -0.61
    eries
    -0.61
    é¾įå
    -0.61
    gard
    -0.60
    mber
    -0.59
    POSITIVE LOGITS
    usra
    0.80
    ...]
    0.74
    acements
    0.74
    aders
    0.74
     ],
    0.68
     EntityItem
    0.67
    =]
    0.67
    ement
    0.66
     Flavoring
    0.65
    ],
    0.65
    Activations Density 0.028%

    No Known Activations