INDEX

Explanations

address-related terms, particularly street names and locations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 3-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_3/width_16k/average_l0_59

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.3.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

ăng

-0.46

цов

-0.44

ifik

-0.42

nest

-0.42

 diikuti

-0.42

CTL

-0.41

neman

-0.41

AVER

-0.41

PLES

-0.41

borg

-0.41

POSITIVE LOGITS

CppMethod

0.98

 EconPapers

0.98

__':

0.96

IsContent

0.90

 дописавши

0.90

__':

0.83

 كومونز

0.82



0.80

StoreMessageInfo

0.80

colhead

0.77

Activations Density 0.162%

address-related terms, particularly street names and locations

No Comments

No Known Activations