INDEX

Explanations

phrases related to self-awareness and introspection

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

sequence "next code"

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

 Consejos

-0.32

 plantes

-0.30

 ketentuan

-0.30

 espalda

-0.30

 chargé

-0.29

 laborales

-0.29

wps

-0.29

 Seit

-0.29

rašymas

-0.29

Seit

-0.28

POSITIVE LOGITS

findpost

0.62

SequentialGroup

0.61

nocache

0.61

Personendaten

0.61

NOPQRST

0.61

windowFixed

0.59

 تضيفلها

0.58

 ddelweddau

0.57

ANTLR

0.56

ۜ

0.56

Activations Density 0.029%

phrases related to self-awareness and introspection

sequence "next code"

No Comments

No Known Activations