INDEX

Explanations

references to "instruction" 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 @ 0-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_0/width_16k/average_l0_105

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

InitStruct

-0.69

 ویکی‌پدیا

-0.68

➟

-0.64

Wart

-0.63

OSI

-0.62

 StatefulWidget

-0.62

mtrl

-0.61

Personensuche

-0.60

 orphanage

-0.58

gym

-0.58

POSITIVE LOGITS

 instruction

1.97

 instructions

1.92

 Instruction

1.88

 Instructions

1.78

Instructions

1.68

instruction

1.64

instructions

1.63

Instruction

1.60

 INSTRUCTIONS

1.57

 INSTRUCTION

1.55

Activations Density 0.132%

references to "instruction" in various contexts

No Comments

No Known Activations