INDEX

Explanations

expressions related to mathematical operations and programming syntax

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

percentages and rates per duration

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

 GetEnumerator

-0.44

 للاسماء

-0.42

 Filho

-0.41

 onCancelled

-0.41

postDelayed

-0.40

FontStyle

-0.39

 ujednoznacz

-0.39

LookAnd

-0.39

期刊论文

-0.38

 zasi

-0.38

POSITIVE LOGITS

parsedMessage

0.48

UnusedPrivate

0.44

inSlope

0.40

#+#

0.40

InputBorder

0.39

enschaften

0.38

其妙

0.37

ánto

0.36

autaire

0.36

SOU

0.35

Activations Density 0.018%

expressions related to mathematical operations and programming syntax

percentages and rates per duration

No Comments

No Known Activations