INDEX

Explanations

terms related to mistakes or incorrectness

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

"wrong" or "incorrect"

np_acts-logits-general · gemini-2.0-flash

accuracy or fault

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

 AssemblyCompany

-0.83

RectangleBorder

-0.74

 GoogleFonts

-0.69

 newBuilder

-0.69

tagHelperRunner

-0.68

LabelTagHelper

-0.65

 behoort

-0.63

 tramonto

-0.62

nthetic

-0.62

StackNavigator

-0.60

POSITIVE LOGITS

 Wrong

1.10

 wrong

1.05

Wrong

0.99

 WRONG

0.98

wrong

0.96

WRONG

0.92

 Incorrect

0.83

Incorrect

0.82

 wrongs

0.79

 incorrect

0.78

Activations Density 0.150%

terms related to mistakes or incorrectness

"wrong" or "incorrect"

accuracy or fault

No Comments

No Known Activations