INDEX

Explanations

phrases indicating dominance or primacy in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

any

-0.68

 something

-0.64

any

-0.63

tagHelper

-0.60

Any

-0.59

something

-0.57

 qualsiasi

-0.57

Next

-0.56

itz

-0.53

 qualquer

-0.52

POSITIVE LOGITS

 Mostly

0.97

Mostly

0.94

+#+#

0.94

 Mainly

0.87

 Primarily

0.86

mostly

0.82

 ExecuteAsync

0.81

 głównie

0.79

 mainly

0.78

mainly

0.78

Activations Density 0.213%

phrases indicating dominance or primacy in various contexts

No Comments

No Known Activations