INDEX

Explanations

code-related terms and structures related to matching and pattern recognition

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

match and score

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-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Portale

-0.55

 Мексичка

-0.43

 Wirtschaft

-0.41

 intptr

-0.41

nullable

-0.40

enumii

-0.39

enumi

-0.38

 ***!

-0.37

awtextra

-0.37

 дописавши

-0.36

POSITIVE LOGITS

 Match

1.03

Match

0.92

match

0.88

 MATCH

0.87

 match

0.87

MATCH

0.86

 Matches

0.73

 matches

0.70

matches

0.69

 matcher

0.67

Activations Density 0.032%

code-related terms and structures related to matching and pattern recognition

match and score

No Comments

No Known Activations