INDEX

Explanations

words related to conclusions or summaries, particularly those that denote a logical progression or implication

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 OTHERWISE

-0.57

というと

-0.55

 otherwise

-0.54

すみません

-0.54

ustu

-0.52

 maxn

-0.52

Okey

-0.50

</i>

-0.50

 communautaire

-0.50

otherwise

-0.49

POSITIVE LOGITS

forth

1.14

 مشين

1.08

invokeLater

0.96

 дописавши

0.93

 مرئيه

0.92

RegistryLite

0.86

 forth

0.82

 &___

0.80

 للاسماء

0.78

styleType

0.77

Activations Density 0.188%

words related to conclusions or summaries, particularly those that denote a logical progression or implication

No Comments

No Known Activations