INDEX

Explanations

references to specific times and durations

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

time expressions

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

ſelf

-0.65

tagHelperRunner

-0.59

 betweenstory

-0.57

OGND

-0.57

 juſ

-0.56

ſelves

-0.51

 queſta

-0.50

 uſe

-0.48

 ſche

-0.48

Diweddarwch

-0.48

POSITIVE LOGITS

onedDateTime

0.53

 midday

0.50

minutes

0.49

 minutes

0.48

 hour

0.46

noon

0.45

 noon

0.45

 afternoon

0.44

 hours

0.43

 mins

0.43

Activations Density 0.064%

references to specific times and durations

time expressions

No Comments

No Known Activations