INDEX

Explanations

occurrences of the word "breakdown" and references to senior individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_65k/average_l0_73

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.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

 uska

-0.64

 Waray

-0.62

 Spisak

-0.59

 kaarangay

-0.55

 Toto

-0.55

 تضيفلها

-0.55

clays

-0.54

 bahay

-0.53

 kaynağından

-0.53

ņas

-0.52

POSITIVE LOGITS

 Senior

0.69

/**

0.69

 senior

0.68

 Seniors

0.67

__*/

0.64

Senior

0.61

senior

0.60

 seniors

0.60

__["

0.60

npmjs

0.58

Activations Density 0.050%

occurrences of the word "breakdown" and references to senior individuals

No Comments

No Known Activations