© 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. 22255
    Prev
    Next
    INDEX
    Explanations

    capitalized acronyms or abbreviations followed by a numerical value probably representing a location or an organization

    oai_token-act-pair · gpt-3.5-turbo

    references to a specific entity or group identified by the letter 'C'

    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
    imony
    -0.70
     wrench
    -0.64
     unborn
    -0.64
     concerned
    -0.59
     shin
    -0.58
     shine
    -0.57
     allotted
    -0.57
     nails
    -0.57
    aterasu
    -0.57
     seeker
    -0.56
    POSITIVE LOGITS
    C
    3.24
    Cs
    2.36
     C
    1.72
    CV
    1.72
    CM
    1.70
    CN
    1.67
    CW
    1.66
    c
    1.65
    CF
    1.65
    CU
    1.63
    Activations Density 0.040%

    No Known Activations