INDEX

Explanations

confirmatory or conditional phrases and qualifiers related to necessity or capability in instruction-based contexts

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-2b-pt-res/layer_2/width_16k/average_l0_141

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

<bos>

-0.95

 anglès

-0.58

 kwamba

-0.55

labus

-0.55

 elettrico

-0.54

 maschile

-0.54

 tarafından

-0.53

 dům

-0.52

 assurent

-0.51

 indipendente

-0.51

POSITIVE LOGITS

AccessorTable

0.83

ंदीखरीदारी

0.74

tvguidetime

0.66

protoimpl

0.65

 letzt

0.65

χρι

0.64

 protoimpl

0.63

 Италијани

0.62

writeFieldEnd

0.60

σιμοποι

0.60

Activations Density 0.629%

confirmatory or conditional phrases and qualifiers related to necessity or capability in instruction-based contexts

No Comments

No Known Activations