INDEX

Explanations

practical advice or tips

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

tip or hint followed by colon

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 hendes

-0.56

 sitten

-0.53

 Infór

-0.51

basicConfig

-0.51

 abstrato

-0.50

protoimpl

-0.49

CloseOperation

-0.48

Clik

-0.48

 zijne

-0.48

 damska

-0.47

POSITIVE LOGITS

tip

0.57

Hint

0.55

 Hint

0.52

Tip

0.48

 estekak

0.48

tip

0.47

Tip

0.46

hint

0.46

bonus

0.45

 bonus

0.45

Activations Density 0.016%

practical advice or tips

tip or hint followed by colon

No Comments

No Known Activations