INDEX

Explanations

relationships between positions and actions

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

preceding "position"

np_acts-logits-general · gemini-2.0-flash

state of position

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 kasarigan

-0.74

]--;

-0.55

setObjectName

-0.53

databind

-0.52

scriptcase

-0.52

biamo

-0.50

OrNil

-0.49

 AttributeSet

-0.48

audiovisuel

-0.47

Atsauces

-0.46

POSITIVE LOGITS

 position

3.37

 Position

2.96

 positions

2.94

position

2.85

 POSITION

2.74

Position

2.69

 Positions

2.54

 positioning

2.44

positions

2.39

 posição

2.38

Activations Density 0.522%

relationships between positions and actions

preceding "position"

state of position

No Comments

No Known Activations