INDEX

Explanations

phrases or words that denote a basis or foundation for claims or arguments

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

 Plummer

-0.74

umato

-0.70

 Flanagan

-0.69

Scu

-0.68

COLUMNS

-0.67

 Greenwich

-0.67

 Displacement

-0.67

 WALT

-0.67

Scu

-0.66

Portale

-0.66

POSITIVE LOGITS

 based

2.34

 Based

2.27

Based

2.23

based

2.16

 BASED

2.08

BASED

2.04

 basé

1.83

 basée

1.68

 basado

1.65

 baseado

1.58

Activations Density 0.132%

phrases or words that denote a basis or foundation for claims or arguments

No Comments

No Known Activations