INDEX

Explanations

steps and processes related to programming or technical instructions

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

Step numbers

np_acts-logits-general · gemini-2.0-flash

numbered steps

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-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 handelt

-0.52

 snippetHide

-0.50

*/;

-0.50

 виправивши

-0.48

kaŭ

-0.46

}=\{

-0.46

 Tampoco

-0.46

Predecesor

-0.45

plays

-0.45

そういえば

-0.45

POSITIVE LOGITS

 step

2.01

 Step

1.82

Step

1.81

step

1.72

 steps

1.70

 Steps

1.57

 STEP

1.56

STEP

1.55

steps

1.42

Steps

1.40

Activations Density 0.253%

steps and processes related to programming or technical instructions

Step numbers

numbered steps

No Comments

No Known Activations