INDEX

Explanations

situations indicating potential problems or abnormalities

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 celui

-0.47

 celle

-0.47

 AssemblyCompany

-0.47

 ones

-0.46

leid

-0.44

τώ

-0.44

Complete

-0.43

Something

-0.43

 something

-0.43

λαμβ

-0.42

POSITIVE LOGITS

 wrong

1.17

 happening

1.15

 amiss

1.08

 happened

1.05

 happen

1.03

 happens

1.00

wrong

0.92

happens

0.88

happened

0.88

happen

0.87

Activations Density 0.270%

situations indicating potential problems or abnormalities

No Comments

No Known Activations