INDEX

Explanations

references to intervention strategies and related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

 Sams

-0.77

 säker

-0.75

Sams

-0.74

rdata

-0.66

 fres

-0.63

 Fres

-0.63

 Pleasure

-0.60

grunt

-0.60

Ama

-0.60

 Trello

-0.59

POSITIVE LOGITS

 intervention

1.22

Intervention

1.16

 Intervention

1.15

 interventions

1.14

intervention

1.12

Interventions

1.12

 Interventions

1.03

 Invo

0.99

 intervene

0.98

 intervenir

0.92

Activations Density 0.157%

references to intervention strategies and related terminology

No Comments

No Known Activations