INDEX

Explanations

combinations of words and phrases that indicate relationships and distributions among features or settings

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

"combinations" or "pairs" of items

np_acts-logits-general · gemini-2.0-flash

combinations and pairs

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k

Configuration

google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.28.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

StoryboardSegue

-0.49

keyColumn

-0.46

المكان

-0.44

kool

-0.40

Greta

-0.40

 BoxDecoration

-0.40

IBOutlet

-0.39

FormTagHelper

-0.39

GRect

-0.38

 okuyayım

-0.38

POSITIVE LOGITS

 combination

2.67

 combinations

2.44

combination

2.28

 Combination

2.25

 combo

2.19

Combination

2.19

 Kombination

2.14

 Combinations

2.13

 combinação

2.09

 combina

1.98

Activations Density 0.758%

combinations of words and phrases that indicate relationships and distributions among features or settings

"combinations" or "pairs" of items

combinations and pairs

No Comments

No Known Activations