INDEX

Explanations

the presence of the word "be" 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 picada

-0.61

 aérienne

-0.60

WriteTagHelper

-0.58

 progressively

-0.58

atalytic

-0.58

År

-0.57

ilever

-0.57

XXIII

-0.57

skraft

-0.57

destruct

-0.57

POSITIVE LOGITS

Not

1.02

(!__

0.83

Not

0.82

NOT

0.78

 Nicht

0.76

就不是

0.76

IsNot

0.70

 necessarily

0.68

 Bukan

0.68

 unlike

0.68

Activations Density 0.219%

the presence of the word "be" in various contexts

No Comments

No Known Activations