INDEX

Explanations

terms related to detection and diagnostic processes

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

years

-0.56

 years

-0.56

 myſelf

-0.53

années

-0.53

 purpoſe

-0.52

isNullOrEmpty

-0.52

 Empres

-0.52

 étrangère

-0.51

Years

-0.51

 année

-0.51

POSITIVE LOGITS

 detection

1.20

 Detect

1.17

 detect

1.16

 Detection

1.10

detect

1.10

 detects

1.01

 detecting

0.99

 detected

0.99

detection

0.98

 detections

0.97

Activations Density 0.037%

terms related to detection and diagnostic processes

No Comments

No Known Activations