INDEX

Explanations

variations and differences across contexts, particularly focusing on heterogeneity and discrepancies in data or practices

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

difference or variation

np_acts-logits-general · gemini-2.0-flash

differ and vary

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 double

-0.51

jedis

-0.49

rokken

-0.49

rlrl

-0.49

ános

-0.48

 perpetuity

-0.47

 doble

-0.47

ведите

-0.47

chossen

-0.46

ppins

-0.46

POSITIVE LOGITS

 depending

0.78

 between

0.73

tvguidetime

0.68

ftagPool

0.67

 قلي

0.62

 wildly

0.62

depending

0.61

AnchorStyles

0.61

 radically

0.60

 tamen

0.60

Activations Density 0.350%

variations and differences across contexts, particularly focusing on heterogeneity and discrepancies in data or practices

difference or variation

differ and vary

No Comments

No Known Activations