INDEX

Explanations

imperative statements directing actions or recommendations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

Comes before an imperative verb

np_acts-logits-general · gemini-2.0-flash

advise or recommend actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Così

-0.35

まさかの

-0.35

 così

-0.33

Suddenly

-0.32

 Suddenly

-0.32

 nouvel

-0.31

 successo

-0.31

blurRadius

-0.31

勝手に

-0.31

 ricevere

-0.31

POSITIVE LOGITS

 remember

0.87

 beware

0.84

 consider

0.83

 always

0.73

 Consider

0.72

 understand

0.72

 jangan

0.71

 Beware

0.69

 keep

0.69

 стара

0.69

Activations Density 0.300%

imperative statements directing actions or recommendations

Comes before an imperative verb

advise or recommend actions

No Comments

No Known Activations