INDEX

Explanations

technical terminology related to communication devices and their functionalities.

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

mentions of technology components and troubleshooting in a technical context.

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

clauses or phrases that introduce or describe a key idea or concept, often using words like "that" or "which" to connect the description to the main subject.

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

"that" followed by a verb or auxiliary verb

np_acts-logits-general · gemini-2.0-flash

that followed by verb

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 defStyle

-0.57

 sqlSession

-0.56

 évidence

-0.54

 utafitiHapana

-0.52

setAccessible

-0.51

est

-0.51

tained

-0.50

 snowing

-0.50

lest

-0.49

 realisation

-0.48

POSITIVE LOGITS

has

1.05

is

1.00

can

0.97

 actually

0.90

 will

0.84

 could

0.80

 either

0.79

 really

0.74

 would

0.72

 cannot