INDEX

Explanations

terms related to reducing size or scale

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

shrink

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

 المعيارى

-0.54

Ney

-0.52

ayla

-0.52

 eventType

-0.52

 boarded

-0.51

 Correspond

-0.51

atoday

-0.50

ceus

-0.50

LookAnd

-0.49

NameInMap

-0.48

POSITIVE LOGITS

shrink

1.16

Shrink

1.13

 shrink

1.10

 Shrink

1.09

 shrinkage

0.98

 shrinks

0.94

 shrunk

0.94

 shrinking

0.92

 shrank

0.88

Shr

0.81

Activations Density 0.007%

terms related to reducing size or scale

shrink

No Comments

No Known Activations