INDEX

Explanations

references to programs or interventions

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

AlterField

-0.80

 médicaux

-0.74

 Oxide

-0.74

etches

-0.73

']

-0.72

 cheeks

-0.71

 Cuthbert

-0.71

+}\

-0.70

 validés

-0.70

Kön

-0.69

POSITIVE LOGITS

 programs

1.88

 program

1.76

 programmes

1.65

 Programs

1.60

 Program

1.58

Programs

1.57

 programme

1.57

PROGRAM

1.55

programs

1.55

Program

1.51

Activations Density 0.035%

references to programs or interventions

No Comments

No Known Activations