INDEX

Explanations

endian and stdint

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

the token "int" in programming contexts, particularly in type declarations like "stdint" or "uint".

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

The neuron strongly fires on tokens denoting byte‐order macros and descriptors—i.e. anything containing “ENDIAN” (e.g. BIG_ENDIAN_DESC, ENDIAN).

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

-2.95

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-2.81

渽

-2.70

</h4>

-2.66

↵↵↵↵↵↵↵↵↵↵

-2.64

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-2.64

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-2.56

↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-2.50

↵↵↵↵↵↵↵↵↵↵↵↵↵

-2.48

↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵

-2.48

POSITIVE LOGITS



3.00

 hırka

2.98

 diki

2.86

 kalem

2.83

 etek

2.73

皛

2.73



2.73

 bunda

2.66

there

2.66

 kollu

2.64

Activations Density 0.006%

endian and stdint

the token "int" in programming contexts, particularly in type declarations like "stdint" or "uint".

The neuron strongly fires on tokens denoting byte‐order macros and descriptors—i.e. anything containing “ENDIAN” (e.g. BIG_ENDIAN_DESC, ENDIAN).

No Comments

No Known Activations

endian and stdint

the token "int" in programming contexts, particularly in type declarations like "stdint" or "uint".

The neuron strongly fires on tokens denoting byte‐order macros and descriptors—i.e. anything containing “ENDIAN” (e.g. BIG_ENDIAN_DESC, ENDIAN).

No Comments

No Known Activations