INDEX

Explanations

terms related to grammatical categories and syntactic structure

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

distinguished by

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

findpost

-0.55

aarrggbb

-0.51

///</

-0.47

complexContent

-0.45

+#+

-0.43

sizeCache

-0.42

WireFormat

-0.42

SharedDtor

-0.42

webElementXpaths

-0.41

Биография

-0.40

POSITIVE LOGITS

0.74

 שוליים

0.57

 مرئيه

0.47

참고

0.45

luents

0.41

Obtain

0.40

 Kampagne

0.40

 Schulen

0.40

ⓘ

0.39

onucleotide

0.38

Activations Density 0.128%

terms related to grammatical categories and syntactic structure

distinguished by

No Comments

No Known Activations