INDEX

Explanations

phrases indicating general references or unspecified subjects

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

everything, nothing, all scope

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

 nakalista

-0.63

 للمعارف

-0.63

 المعيارى

-0.61

ScopeManager

-0.60

URLException

-0.60

 يتيمه

-0.60

 queſta

-0.59

__*/

-0.57

IUrlHelper

-0.57

ImageContext

-0.56

POSITIVE LOGITS

0.30

 casais

0.29

fit

0.29

 thread

0.29

 Offisielt

0.29

anglès

0.28

 generation

0.28

 unit

0.28

 campagnes

0.28

Activations Density 0.012%

phrases indicating general references or unspecified subjects

everything, nothing, all scope

No Comments

No Known Activations