INDEX

Explanations

specific names, terms, or concepts in technical or scientific contexts

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

 ویکی‌پدی

-0.62

 Meksiku

-0.59

OGND

-0.59

Tikang

-0.57

 Италијани

-0.51

EndGlobalSection

-0.51

 podjela

-0.51

/**

-0.50

 незавершена

-0.49

AddWithValue

-0.48

POSITIVE LOGITS

 <=",

0.47

WriteTagHelper

0.38

mercato

0.35

írus

0.35

ostante

0.35

zegovina

0.35

utilisons

0.34

remadura

0.34

vay

0.34

RuleContext

0.32

Activations Density 0.050%

specific names, terms, or concepts in technical or scientific contexts

No Comments

No Known Activations