INDEX

Explanations

mentions of locations and organizations related to healthcare

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

Followed by "the" or specific letters

np_acts-logits-general · gemini-2.0-flash

Sussex place names

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

derry

-0.60

 ujednoznacz

-0.54

 esternos

-0.52

 Oberlin

-0.50

RenderAtEndOf

-0.49

 Harlem

-0.48

ulipas

-0.48

Zar

-0.48

 HttpNotFound

-0.46

 blom

-0.46

POSITIVE LOGITS

 Sussex

1.22

 Brighton

0.91

 Lewes

0.86

Brighton

0.83

 Crawley

0.82

 Hove

0.81

Sus

0.74

 Arundel

0.73

 Hastings

0.69

 Gatwick

0.68

Activations Density 0.155%

mentions of locations and organizations related to healthcare

Followed by "the" or specific letters

Sussex place names

No Comments

No Known Activations