INDEX

Explanations

references to emotions or feelings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

ⓧ

-0.63

 houſe

-0.52

 Dubai

-0.44

PageFactory

-0.43

 Houſe

-0.43

multer

-0.42

 entrer

-0.41

Autowired

-0.41

 door

-0.40

Far

-0.40

POSITIVE LOGITS

 Sentiment

1.20

Sentiment

1.16

 sentiment

1.13

sentiment

1.05

 Senti

0.96

 sentiments

0.92

Senti

0.89

 sentimental

0.81

senti

0.79

entiment

0.75

Activations Density 0.005%

references to emotions or feelings

No Comments

No Known Activations