INDEX

Explanations

liquid extraction, in vitro

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

special formatting characters and symbols often found in technical documents or code.

oai_token-act-pair · claude-3-7-sonnet-20250219 Triggered by @neilrathi

The neuron fires strongly on numeric tokens—years, version numbers, decimal measurements, gene‐counts, etc.—i.e. any standalone numbers or number‐formatted strings.

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

so

-2.06

 enfans

-2.02

 prêtres

-1.88

罫

-1.82

 chrétiens

-1.81

 espirituales

-1.80

匾

-1.76

 religieuses

-1.73

 Theologie

-1.70

 sacerdotes

-1.69

POSITIVE LOGITS

ֶּ

2.03

 مباراة

1.95

 сообщили

1.88

ᾰ

1.80

чя

1.80

vább

1.74

 torsdag

1.71

 indicó

1.67

1.66

 Sabtu

1.64

Activations Density 0.003%

liquid extraction, in vitro

special formatting characters and symbols often found in technical documents or code.

The neuron fires strongly on numeric tokens—years, version numbers, decimal measurements, gene‐counts, etc.—i.e. any standalone numbers or number‐formatted strings.

No Comments

No Known Activations

liquid extraction, in vitro

special formatting characters and symbols often found in technical documents or code.

The neuron fires strongly on numeric tokens—years, version numbers, decimal measurements, gene‐counts, etc.—i.e. any standalone numbers or number‐formatted strings.

No Comments

No Known Activations