INDEX

Explanations

specific references to locations and transportation information.

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

text related to bus schedules, transportation options, and geographical locations.

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

place names and geographic locations, particularly those in Finland and other Nordic countries.

oai_token-act-pair · claude-3-5-sonnet-20240620 Triggered by @bot

Finnish locations or organizations

np_acts-logits-general · gemini-2.0-flash

Finnish locations and nationalities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.20.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

 vastaan

-0.68

 avulla

-0.66

 kanssa

-0.64

 vuotta

-0.62

 kautta

-0.61

 jälkeen

-0.61

 aikana

-0.60

 jäsen

-0.59

 päivä

-0.59

 ruok

-0.58

POSITIVE LOGITS

 himo

0.80

évaluateur

0.79

 Finnish

0.79

tagHelperRunner

0.77

 Turku

0.74

 sauna

0.72

 Helsinki

0.71

 Finns

0.71

 Estonian

0.69

RenderAtEndOf

0.68

Activations Density 0.127%