INDEX

Explanations

phrases indicating uncertainty or qualifications in statements

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

some certain noun

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

 tarko

-0.47

 itſelf

-0.43

 thuận

-0.43

 hyö

-0.41

 BRARY

-0.41

 proprement

-0.40

 drewna

-0.39

 jäl

-0.39

rsiniz

-0.39

本体

-0.38

POSITIVE LOGITS

ThroughAttribute

0.52

 niektórych

0.48

 egyes

0.48

某些

0.46

 certain

0.46

########.

0.45

RegressionTest

0.45

 بعضی

0.45

certain

0.45

 sommige

0.45

Activations Density 0.068%

phrases indicating uncertainty or qualifications in statements

some certain noun

No Comments

No Known Activations