INDEX

Explanations

phrases indicating changes or unexpected events

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

Preceding words of "when"

np_acts-logits-general · gemini-2.0-flash

when conjunction

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

WriteBarrier

-0.60

tvguidetime

-0.58

MLLoader

-0.56

 jScrollPane

-0.56

 endblock

-0.55

DockStyle

-0.55

Tembelea

-0.54

titleMargin

-0.50

AISSEE

-0.49

addPreferredGap

-0.47

POSITIVE LOGITS

 when

3.27

 quando

2.66

 cuando

2.61

when

2.58

 когда

2.44

 When

2.17

 WHEN

2.16

 lorsque

2.14

When

2.13

 när

2.11

Activations Density 1.297%

phrases indicating changes or unexpected events

Preceding words of "when"

when conjunction

No Comments

No Known Activations