INDEX

Explanations

references to organizations and their operational details

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

used to or often

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

currently

-0.77

Currently

-0.74

 Currently

-0.73

 currently

-0.71

 queſta

-0.65

setVerticalGroup

-0.63

 presently

-0.60

 derzeit

-0.60

RenderAtEndOf

-0.60

 momentan

-0.59

POSITIVE LOGITS

 lived

0.51

was

0.50

 often

0.48

 gustaba

0.47

 liked

0.43

 frecuente

0.42

 thrived

0.42

 gave

0.42

 always

0.41

 loved

0.41

Activations Density 0.249%

references to organizations and their operational details

used to or often

No Comments

No Known Activations