INDEX

Explanations

SQL JOIN clauses

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

SQL JOIN clauses, particularly INNER JOIN and LEFT JOIN keywords.

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

The neuron detects SQL join clauses by firing strongly on JOIN keywords (e.g. INNER JOIN, LEFT JOIN, OUTER JOIN).

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

and

-2.33

–

-2.05

 что

-1.88

 nadat

-1.84

↵↵

-1.84

“

-1.77

 volgt

-1.77

</h3>

-1.76

 vollständige

-1.65

if

-1.60

POSITIVE LOGITS

錞

2.13

 hırka

1.95

 ۳۰

1.93

 marié

1.85

偁

1.83

卨

1.77

 incrí

1.76

Karakteristik

1.75

 camiset

1.74

嫆

1.73

Activations Density 0.008%

SQL JOIN clauses

SQL JOIN clauses, particularly INNER JOIN and LEFT JOIN keywords.

The neuron detects SQL join clauses by firing strongly on JOIN keywords (e.g. INNER JOIN, LEFT JOIN, OUTER JOIN).

No Comments

No Known Activations

SQL JOIN clauses

SQL JOIN clauses, particularly INNER JOIN and LEFT JOIN keywords.

The neuron detects SQL join clauses by firing strongly on JOIN keywords (e.g. INNER JOIN, LEFT JOIN, OUTER JOIN).

No Comments

No Known Activations