INDEX

Explanations

references to heat and temperature

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

heat + technical terms

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

 للاسماء

-0.74

endcsname

-0.64

carpa

-0.63

 propOrder

-0.63

cytic

-0.59

zweig

-0.57

 betweenstory

-0.57

 Milli

-0.57

enumii

-0.56

 ErrIntOverflow

-0.55

POSITIVE LOGITS

Heat

1.22

 Heat

1.16

heat

1.06

 heat

1.03

 HEAT

1.02

HEAT

0.99

heatmap

0.82

 Hitze

0.79

 Wärme

0.77

 Heaton

0.74

Activations Density 0.013%

references to heat and temperature

heat + technical terms

No Comments

No Known Activations