INDEX

Explanations

specific technical terms and numerical values related to computing and scientific contexts

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

votto, o'tool,acterial

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

BufferException

-0.85

enderror

-0.84

 ujednoznacz

-0.78

ScopeManager

-0.76

IsContent

-0.76

KommentareTeilen

-0.74

 كومونز

-0.74

Diwedd

-0.73

postIndex

-0.71

帖最后由

-0.71

POSITIVE LOGITS

 chêne

0.27

 marbre

0.27

 Jerusalén

0.26

 exitoso

0.26

ducir

0.25

 Nación

0.25

 usted

0.24

 Hause

0.24

REPEAT

0.23

writel

0.23

Activations Density 0.036%

specific technical terms and numerical values related to computing and scientific contexts

votto, o'tool,acterial

No Comments

No Known Activations