INDEX

Explanations

predictions or assumptions made by the speaker

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 9-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_9/width_16k/average_l0_73

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.9.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

 betweenstory

-0.68

 objectForKey

-0.63

Aholisi

-0.63

tigny

-0.61

 nakalista

-0.60

mployment

-0.60

setzungen

-0.60

++++++++++++++++

-0.55

ronique

-0.54

Abp

-0.54

POSITIVE LOGITS

guess

0.93

 guess

0.81

Guess

0.80

 Guess

0.80

 maybe

0.77

__()

0.68

GroupLayout

0.66

 perhaps

0.65

γον

0.64

 insuffisamment

0.64

Activations Density 0.058%

predictions or assumptions made by the speaker

No Comments

No Known Activations