INDEX

Explanations

phrases indicating the passage of time or recent actions

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

now followed by completed action

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

昔から

-0.36

 gilet

-0.35

 diarr

-0.35

FixedSize

-0.33

 esqu

-0.32

essica

-0.32

 tradition

-0.31

irited

-0.31

前から

-0.31

 MemoryStream

-0.31

POSITIVE LOGITS

IntoConstraints

0.70



0.57

 AssemblyTitle

0.56

 насељу

0.53

 שוליים

0.51

RegressionTest

0.49

 aarde

0.48

 verfügen

0.47

liferay

0.47

帖最后由

0.47

Activations Density 0.007%

phrases indicating the passage of time or recent actions

now followed by completed action

No Comments

No Known Activations