INDEX

Explanations

transitional phrases indicating summaries or conclusions

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

"So" followed by a comma or a question

np_acts-logits-general · gemini-2.0-flash

so followed by new ideas

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

ſelves

-1.04

ſelf

-1.03

 Majefty

-0.95

 Shakspeare

-0.90

 Houſe

-0.86

Efq

-0.85

 myſelf

-0.84

SBATCH

-0.81

 entanto

-0.80

 neceff

-0.80

POSITIVE LOGITS

 what

0.86

oner

0.79

 basically

0.77

if

0.76

So

0.73

yes

0.73

So

0.71

 yeah

0.65

how

0.65

Activations Density 0.087%

transitional phrases indicating summaries or conclusions

"So" followed by a comma or a question

so followed by new ideas

No Comments

No Known Activations