INDEX

Explanations

instances of gaps or disparities in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 5-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_5/width_16k/average_l0_68

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.5.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

Shou

-0.90

 dioses

-0.77

 principes

-0.68

 McCarty

-0.66

 carros

-0.64

 genitori

-0.64

 Roscoe

-0.63

 torchvision

-0.63

Boas

-0.63

 Cartwright

-0.62

POSITIVE LOGITS

gap

2.82

Gap

2.58

gap

2.47

 gaps

2.44

Gap

2.36

 Gaps

2.35

gaps

2.06

GAP

1.98

GAP

1.66

 chasm

1.25

Activations Density 0.103%

instances of gaps or disparities in various contexts

No Comments

No Known Activations