INDEX

Explanations

superlative expressions that denote quality and ranking

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

quantifying superlatives

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

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

RenderAtEndOf

-0.73

 Wicidata

-0.61

MigrationBuilder

-0.58

 FetchType

-0.56

 purpoſe

-0.55

 ſta

-0.54

sidemargin

-0.54

createCanvas

-0.54

webElementXpaths

-0.54



-0.53

POSITIVE LOGITS

 Besten

0.42

 best

0.41

 decorazione

0.41

 SELECTION

0.40

 popularity

0.39

 popularidad

0.39

 najbolj

0.39

 Best

0.38

 favourite

0.38

 nejlepší

0.38

Activations Density 0.039%

superlative expressions that denote quality and ranking

quantifying superlatives

No Comments

No Known Activations

superlative expressions that denote quality and ranking

quantifying superlatives

No Comments

No Known Activations