INDEX

Explanations

references to organizations or entities related to infrastructure

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

Followed by tokens like "NH" and "infrastructure"

np_acts-logits-general · gemini-2.0-flash

NHTSA, NH4+, infrastructure, Russian words

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

•Gemma-2-2B-L20-Narrative-Cluster

Natural Language (relatively inscrutable, but code presence is very strong in negative logits)

No Comments

Negative Logits



-0.96

noinspection

-0.80

 ArrayAdapter

-0.79

endpush

-0.79

ceutical

-0.78

 Réponses

-0.78

TextWatcher

-0.78

ErrorListener

-0.76

ADVERTISEMENT

-0.74

 springfox

-0.74

POSITIVE LOGITS

0.80

0.68

0.67

ます

0.66

い

0.58

躇

0.58

 sare

0.57

 snapping

0.56

 socc

0.54

"):

0.52

Activations Density 0.697%

references to organizations or entities related to infrastructure

Followed by tokens like "NH" and "infrastructure"

NHTSA, NH4+, infrastructure, Russian words

No Comments

No Known Activations