INDEX

Explanations

phrases related to emotional or reflective experiences

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

Sentences followed by conjunctions or continuations

np_acts-logits-general · gemini-2.0-flash

possibilities and questions

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

(;;)

-0.78

saraba

-0.64

клопе

-0.63

glGen

-0.63

ungeon

-0.61

 wikipagina

-0.61

？】

-0.60

Cited

-0.60

THISDAY

-0.59

 />";

-0.59

POSITIVE LOGITS

 kasarigan

0.63

 چقدر

0.57

 impost

0.56

Damn

0.51

Thinking

0.51

damn

0.50

 maybe

0.50

 thought

0.49

ImageContext

0.49

should

0.48

Activations Density 0.107%

phrases related to emotional or reflective experiences

Sentences followed by conjunctions or continuations

possibilities and questions

No Comments

No Known Activations