INDEX

Explanations

terms related to duration, particularly focusing on the long-term versus short-term implications

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

"long" followed by "run" or "term"

np_acts-logits-general · gemini-2.0-flash

long run, long term

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 today

-0.47

WriteLiteral

-0.45

Pyx

-0.45

specialchars

-0.40

gtest

-0.39

ez

-0.39

Referanser

-0.38

Braz

-0.38

solid

-0.38

Ver

-0.38

POSITIVE LOGITS

 term

0.94

term

0.91

 <>",

0.78

Term

0.75

 Term

0.74

 TERM

0.74

 betweenstory

0.71

 OFDb

0.71

 éché

0.71

 oprot

0.70

Activations Density 0.099%

terms related to duration, particularly focusing on the long-term versus short-term implications

"long" followed by "run" or "term"

long run, long term

No Comments

No Known Activations