INDEX

Explanations

negative references to actions not taken or failures to act

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

 ब्रेकडाउन

-0.43

tanleria

-0.38

 iconTwitter

-0.35

defineProperty

-0.33

etCode

-0.30

OSSARY

-0.30

/**

-0.30

 AssemblyTitle

-0.29

ひとつ

-0.28

WebElementEntity

-0.28

POSITIVE LOGITS

.*")]

0.69

 հղումներ

0.67

GraphicsUnit

0.67

tagHelperRunner

0.63

WriteTagHelper

0.58

XtraGrid

0.58

transQ

0.57

 EconPapers

0.57

 kasarigan

0.57

⏟

0.54

Activations Density 0.358%

negative references to actions not taken or failures to act

No Comments

No Known Activations