INDEX

Explanations

intensities of emotional or tense situations

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

Sentences that express a reason or explanation

np_acts-logits-general · gemini-2.0-flash

mystery and unexplained

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-16k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

bim

-0.43

 calor

-0.42

Alte

-0.41

 Alte

-0.41

 Acti

-0.39

ivir

-0.39

OCH

-0.39

UnusedPrivate

-0.38

 rés

-0.38

 Anchor

-0.38

POSITIVE LOGITS

 '\\;'

0.72

 tajem

0.53

transQ

0.53

 mysterious

0.52

 parezca

0.51

 mystérie

0.50

 misterioso

0.50

 misteriosa

0.49

tvguidetime

0.49

uvwxyz

0.48

Activations Density 0.737%

intensities of emotional or tense situations

Sentences that express a reason or explanation

mystery and unexplained

No Comments

No Known Activations