INDEX

Explanations

references to legal cases and court decisions

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

versus in legal citations

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

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 avoient

-0.52

Tikang

-0.50

 næste

-0.49

fø

-0.48

 nødven

-0.48

 zijne

-0.47

她們

-0.47

 faudrait

-0.45

 būtų

-0.44

UnusedPrivate

-0.44

POSITIVE LOGITS

0.79

0.66

vs

0.56

 valueForKey

0.52

HostException

0.52

versus

0.51

mappedBy

0.50

帖最后由

0.50

Activations Density 0.065%

references to legal cases and court decisions

versus in legal citations

No Comments

No Known Activations