INDEX

Explanations

geographical and political boundary references

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

geographical borders

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

 robust

-0.44

 mobili

-0.44

大家的

-0.41

 positioning

-0.41

Redirection

-0.40

นาย

-0.40

 filter

-0.40

 filtered

-0.39

 Robust

-0.38

arini

-0.38

POSITIVE LOGITS

sizeCache

0.60

 invokingState

0.58

 borders

0.55

EDEFAULT

0.54

 ujednoznacz

0.54

 border

0.53

Autoritní

0.51

CppMethod

0.51

 فريبيس

0.51

HostException

0.48

Activations Density 0.012%

geographical and political boundary references

geographical borders

No Comments

No Known Activations

geographical and political boundary references

geographical borders

No Comments

No Known Activations