INDEX

Explanations

specific numerical values and sequences

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

list items, numbers, and punctuation

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

tagHelperRunner

-0.79

enderror

-0.62

WriteLiteral

-0.61

AndEndTag

-0.60

migrationBuilder

-0.60

AsUp

-0.59

 ProtoMessage

-0.58

]=>

-0.57

@"/

-0.55

 oprot

-0.55

POSITIVE LOGITS

RTGC

0.39

\{\\

0.33

相同

0.33

 dilution

0.30

spunkt

0.29

 nonatomic

0.29

 loco

0.28

 Hollow

0.27

 toxin

0.27

fony

0.27

Activations Density 0.026%

specific numerical values and sequences

list items, numbers, and punctuation

No Comments

No Known Activations