INDEX

Explanations

concepts related to teaching and learning activities

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

actions relating to specific things

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

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

Билгалдахарш

-0.92

RegressionTest

-0.91

ChildScrollView

-0.72

rrggbb

-0.70

LEncoder

-0.70

Diweddarwch

-0.69

fillType

-0.63

findpost

-0.61

aarrggbb

-0.60

 esternos

-0.58

POSITIVE LOGITS

/\.

0.61

 freely

0.54

 them

0.51

 privately

0.51

the

0.51

 свои

0.50

 aggressively

0.50

locate

0.50

 anything

0.49

the

0.49

Activations Density 0.191%

concepts related to teaching and learning activities

actions relating to specific things

No Comments

No Known Activations