INDEX

Explanations

instances of specific numerical and time-related information

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

timestamps

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

 EconPapers

-0.49

 שוליים

-0.47

 TextInputType

-0.47

rolla

-0.47

tabl

-0.45

 early

-0.43

EnableWeb

-0.42

✭✭

-0.42

참고

-0.41

 roll

-0.40

POSITIVE LOGITS

 '\\;'

0.59

WebServlet

0.43

BeginInit

0.39

 Италијани

0.38

cassert

0.38

 NSCoder

0.38

BnF

0.38

 yapmak

0.37

ształ

0.36

tizens

0.36

Activations Density 0.376%

instances of specific numerical and time-related information

timestamps

No Comments

No Known Activations