INDEX

Explanations

the word "more" and its variations indicating increase or comparison

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

Follows "more" and is often an adverb

np_acts-logits-general · gemini-2.0-flash

More importantly, specifically

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

principalColumn

-0.69

 AssemblyCulture

-0.66

utilisons

-0.65

Diweddarwch

-0.65

antMatchers

-0.61

principalTable

-0.61

ftagPool

-0.61

 ComVisible

-0.60

ایج

-0.60

terraform

-0.59

POSITIVE LOGITS

 importantly

1.20

 specifically

0.94

 Specifically

0.79

Specifically

0.79

 Importantly

0.77

 recently

0.72

 interestingly

0.71

 específicamente

0.70

specifically

0.69

 surprisingly

0.68

Activations Density 0.130%

the word "more" and its variations indicating increase or comparison

Follows "more" and is often an adverb

More importantly, specifically

No Comments

No Known Activations