INDEX

Explanations

references to tanks 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 @ 2-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_2/width_16k/average_l0_141

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.2.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

ządu

-0.79

 médec

-0.73

 picioare

-0.66

Hentet

-0.64

istiques

-0.63

umque

-0.62

 Quince

-0.62

CONSIN

-0.61

 dreapta

-0.61

yntaxException

-0.61

POSITIVE LOGITS

BufferException

0.83

 tank

0.78

tank

0.78

 <<<<<<<<<<<<<<

0.78

hook

0.75

 tanks

0.74

 mounted

0.74

 Tank

0.72

</s>

0.70

 expansion

0.70

Activations Density 0.081%

references to tanks in various contexts

No Comments

No Known Activations