INDEX

Explanations

phrases and terms related to official roles and contexts

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

in his capacity or context

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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



-0.42

Spoljašnje

-0.34

GTCX

-0.34

RULES

-0.32

 UIKit

-0.32

regler

-0.32

rubin

-0.32

found

-0.31

stored

-0.31

 dest

-0.30

POSITIVE LOGITS

 contexts

0.78

 context

0.66

 Zusammenhang

0.61

 contextos

0.60

 contexte

0.60

 Context

0.59

 контек

0.59

 Kontext

0.57

HandlerContext

0.57

contexts

0.56

Activations Density 0.076%

phrases and terms related to official roles and contexts

in his capacity or context

No Comments

No Known Activations