INDEX

Explanations

expressions of opinion or belief about key details

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

to be fair

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

 OFDb

-0.44

 otomatig

-0.44

 Италијани

-0.40

-0.38

PerformLayout

-0.38

 RISERV

-0.37

SBATCH

-0.36

 actifs

-0.35

 мәкал

-0.35

 käytet

-0.35

POSITIVE LOGITS

gebob

0.59

Avo

0.55

justed

0.52

Addo

0.52

ABUL

0.52

 misunder

0.51

arote

0.51

 Anson

0.50

uvwxyz

0.49

Alo

0.49

Activations Density 0.071%

expressions of opinion or belief about key details

to be fair

No Comments

No Known Activations