INDEX

Explanations

phrases indicating obligations or recommendations

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

expressions of obligation and recommendation

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

 nifer

-0.48

Abit

-0.46

ficulties

-0.46

*-*-

-0.46

Transcripción

-0.46

Garonne

-0.46

tainers

-0.45

iress

-0.45

ioen

-0.44

*~*~

-0.44

POSITIVE LOGITS

 should

0.52

 Should

0.46

 always

0.45

Should

0.43

 propOrder

0.42

AndroidJUnit

0.41

should

0.41

 seharusnya

0.39

 SHOULD

0.39

不应该

0.38

Activations Density 0.037%

phrases indicating obligations or recommendations

expressions of obligation and recommendation

No Comments

No Known Activations