INDEX

Explanations

code-related structures and concepts within a programming context

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

changes in order or position

np_acts-logits-general · gemini-2.0-flash

swapping, reversing, or exchanging

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

findpost

-0.57

 vidare

-0.50

addHandler

-0.50

ngrx

-0.49

 voorbij

-0.49

 authorised

-0.47

 approximations

-0.46

Referències

-0.46

 Kenyon

-0.46

原始内容存档于

-0.45

POSITIVE LOGITS

 swap

1.94

 reversed

1.82

 swapped

1.81

 Swap

1.79

 swapping

1.75

swap

1.71

 reversal

1.71

 reverse

1.63

 reversing

1.58

Swap

1.58

Activations Density 0.422%

code-related structures and concepts within a programming context

changes in order or position

swapping, reversing, or exchanging

No Comments

No Known Activations