INDEX

Explanations

conditional statements and discussions of possibilities

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

that indicate doubt or reluctance

np_acts-logits-general · gemini-2.0-flash

conjunctions and discourse markers

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

vanced

-0.40

badlogic

-0.36

 HYPER

-0.35

 bienvenida

-0.34

 moderation

-0.33

 متعلقه

-0.33

<bos>

-0.32

Referències

-0.32

ionic

-0.30

CodeDom

-0.30

POSITIVE LOGITS

bukkit

0.55

 bezeichneter

0.52

 nevertheless

0.52

 NSCoder

0.50

Autoritní

0.49

AISSEE

0.48

 reluctantly

0.48

theless

0.48

 nonetheless

0.47

yyr

0.47

Activations Density 0.408%

conditional statements and discussions of possibilities

that indicate doubt or reluctance

conjunctions and discourse markers

No Comments

No Known Activations