INDEX

Explanations

negations and conditional phrases related to availability or recommendations

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

Token after "wouldn't"

np_acts-logits-general · gemini-2.0-flash

wouldn't/won't followed by verb

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

 meant

-0.62

 supposed

-0.56

 '\\;'

-0.55

 also

-0.54

not

-0.54

 Préférences

-0.52

addGap

-0.51

 McQu

-0.50

 وتسجيلات

-0.49

ifoli

-0.49

POSITIVE LOGITS

toler

0.65

RenderAtEndOf

0.64

TouchListener

0.61

 countenance

0.61

 tolerated

0.61

 budge

0.60

 tolerate

0.60

 createState

0.59

 stoop

0.59

 SURV

0.56

Activations Density 0.145%

negations and conditional phrases related to availability or recommendations

Token after "wouldn't"

wouldn't/won't followed by verb

No Comments

No Known Activations