INDEX

Explanations

Bytes and Appellants

np_acts-logits-general · gemini-2.5-flash-lite

the word "Appellants" in legal documents.

oai_token-act-pair · claude-4-5-sonnet Triggered by @bfgcn7frf5

The neuron consistently fires on the literal token “Bytes” (e.g. class or variable names referring to bytes) in code snippets.

oai_token-act-pair · o4-mini Triggered by @jyhe0408

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

琿

-2.45

鴆

-2.36

er

-2.31

櫪

-2.25

楫

-2.23

務省

-2.19

 knows

-2.19

()

-2.09

屛

-2.09

ed

-2.08

POSITIVE LOGITS

2.39

他们的

2.30

2.28

we

2.20

杁

2.19

2.13

完结

2.11

を使用した

2.09

艽

2.08

end

2.06

Activations Density 0.004%

Bytes and Appellants

the word "Appellants" in legal documents.

The neuron consistently fires on the literal token “Bytes” (e.g. class or variable names referring to bytes) in code snippets.

No Comments

No Known Activations

Bytes and Appellants

the word "Appellants" in legal documents.

The neuron consistently fires on the literal token “Bytes” (e.g. class or variable names referring to bytes) in code snippets.

No Comments

No Known Activations