INDEX

Explanations

code-related syntax elements, particularly focusing on method declarations and structure

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

user turn definitions

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

parsedMessage

-0.71

transQ

-0.70

RegressionTest

-0.69

UserScript

-0.69

 '\\;'

-0.61

 betweenstory

-0.60

Бахар

-0.58

CodedInputStream

-0.57

PerformLayout

-0.57

Rujuakan

-0.57

POSITIVE LOGITS

Synonymes

0.35

 adopción

0.35

 vectorielles

0.32

estación

0.31

 interpretación

0.29

ByType

0.29

 Wich

0.28

 Bundesrepublik

0.28

appartamento

0.28

 Noviembre

0.28

Activations Density 0.000%

code-related syntax elements, particularly focusing on method declarations and structure

user turn definitions

No Comments

No Known Activations