INDEX

Explanations

expressions of confusion or uncertainty

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

"I don't" or "I do not"

np_acts-logits-general · gemini-2.0-flash

don't know or care

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 lenker

-0.70

azen

-0.60

transQ

-0.58

MSR

-0.58

 parseFrom

-0.58

rtz

-0.58

 calendriers

-0.58

uncher

-0.57

+][

-0.57

prüche

-0.57

POSITIVE LOGITS

 blame

0.58

AnimationFrame

0.57

 sure

0.56

RemindMe

0.53

 believe

0.53

 know

0.52

 really

0.51

HasForeignKey

0.51

 fancy

0.49

enumi

0.49

Activations Density 0.146%

expressions of confusion or uncertainty

"I don't" or "I do not"

don't know or care

No Comments

No Known Activations