© 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. 16385
    Prev
    Next
    INDEX
    Explanations

    words related to criminal activities, particularly focusing on terms related to murder

    oai_token-act-pair · gpt-3.5-turbo
    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
    stem
    -0.71
     Premium
    -0.70
    API
    -0.69
    ost
    -0.67
    whe
    -0.67
     Appl
    -0.67
    Tokens
    -0.67
    ota
    -0.65
     ECO
    -0.64
    PAC
    -0.63
    POSITIVE LOGITS
     murder
    3.48
     murders
    2.61
     Murder
    2.56
     homicide
    2.38
     murdering
    2.18
     murderer
    2.12
     assassination
    2.06
     murderers
    2.05
     manslaughter
    2.03
     killings
    2.00
    Activations Density 0.028%

    No Known Activations