INDEX

Explanations

conditional phrases and alternatives

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

followed by a preposition

np_acts-logits-general · gemini-2.0-flash

albeit qualifier words like only, not, without

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.18.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

 wireType

-0.55

 createSlice

-0.53

TagHelper

-0.51

Autowired

-0.49

'{@

-0.49

்த்த

-0.48

{§

-0.47

tagHelperRunner

-0.47

 EconPapers

-0.46

 собой

-0.45

POSITIVE LOGITS

via

1.30

 without

1.26

 through

1.18

 with

1.15

without

1.01

through

0.96

 WITHOUT

0.95

via

0.94

by

0.93

 tramite

0.89

Activations Density 0.548%

conditional phrases and alternatives

followed by a preposition

albeit qualifier words like only, not, without

No Comments

No Known Activations