INDEX

Explanations

terms that indicate improvement or enhancement in various contexts

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

followed by "better"

np_acts-logits-general · gemini-2.0-flash

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.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

 étoient

-0.64

 avoient

-0.64

 būt

-0.60

 zijne

-0.58

 strictement

-0.55

verwijspagina

-0.54

 mijne

-0.53

 toiminta

-0.52

ArrowToggle

-0.52

 unbedingt

-0.52

POSITIVE LOGITS

tagHelperRunner

0.53

 integrate

0.49

 мәкал

0.45

heartedly

0.44

 evaluate

0.43

quely

0.43

 encode

0.42

 decode

0.42

 publici

0.41

 penetrate

0.41

Activations Density 0.398%

terms that indicate improvement or enhancement in various contexts

followed by "better"

No Comments

No Known Activations