INDEX

Explanations

phrases referring to specific conditions or situations

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

state or finding of condition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.16.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

揄

-0.87

 NDEBUG

-0.82

🏻

-0.77

Bam

-0.74

GRAFIA

-0.71

 FormBuilder

-0.71

 Somer

-0.70

restle

-0.70

 VIAF

-0.70

 astray

-0.69

POSITIVE LOGITS

 conditions

1.45

 Conditions

1.30

conditions

1.26

Conditions

1.22

Condition

1.22

 condition

1.18

 Condition

1.15

 CONDITIONS

1.10

condition

1.08

 CONDITION

1.02

Activations Density 0.042%

phrases referring to specific conditions or situations

state or finding of condition

No Comments

No Known Activations