INDEX

Explanations

references to official documents and attachments in written communication

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

"Attached" or attachment-related language

np_acts-logits-general · gemini-2.0-flash

attached or appended material

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

OCCURRED

-0.60

elemField

-0.56

 autorytatywna

-0.55

setVerticalGroup

-0.52

唆

-0.52

eleste

-0.51

typeorm

-0.51

конец

-0.51

andExpect

-0.50

 Polres

-0.50

POSITIVE LOGITS

 attached

2.48

Attached

2.22

 Attached

2.19

attached

2.18

 attach

2.06

 attachment

2.02

 attachments

1.97

 attaching

1.88

 Attach

1.83

 Attachment

1.79

Activations Density 0.779%

references to official documents and attachments in written communication

"Attached" or attachment-related language

attached or appended material

No Comments

No Known Activations