INDEX

Explanations

late or earlier events and diagnoses

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

phrases about timing, especially situations where something happens "late" or when time-sensitive processes occur.

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

The neuron strongly activates on the negative adverb “no” (and to a lesser extent on other negative/negation-related words).

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

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_34/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

越多

-0.88

Short

-0.83

initial

-0.81

 больше

-0.80

不大

-0.79

 corto

-0.77

不多

-0.76

*$

-0.76

short

-0.76

短

-0.75

POSITIVE LOGITS

 later

2.27

 late

2.17

 latest

1.87

 earlier

1.87

 relatively

1.84

 very

1.67

 much

1.62

 earliest

1.60

较

1.57

later

1.54

Activations Density 0.073%

late or earlier events and diagnoses

phrases about timing, especially situations where something happens "late" or when time-sensitive processes occur.

The neuron strongly activates on the negative adverb “no” (and to a lesser extent on other negative/negation-related words).

No Comments

No Known Activations

late or earlier events and diagnoses

phrases about timing, especially situations where something happens "late" or when time-sensitive processes occur.

The neuron strongly activates on the negative adverb “no” (and to a lesser extent on other negative/negation-related words).

No Comments

No Known Activations