INDEX

Explanations

programming function definitions and references

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

variable assignment value

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 arşivlendi

-0.88

Diweddarwch

-0.86

Tikang

-0.77

+#+#

-0.76

 للاسماء

-0.75

 الحره

-0.75

 препратки

-0.74

 HasFactory

-0.72

Tembelea

-0.69

ContentAsync

-0.69

POSITIVE LOGITS

value

0.66

 value

0.59

Value

0.44

么

0.38

 Value

0.37

VALUE

0.37

 fapt

0.37

 waarde

0.36

“

0.35

 Vermögen

0.35

Activations Density 0.003%

programming function definitions and references

variable assignment value

No Comments

No Known Activations