INDEX

Explanations

phrases related to escalating conflicts and tensions between countries or groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

cise

-0.84

prints

-0.77

sole

-0.70

Interstitial

-0.69

Chance

-0.67

icle

-0.66

itte

-0.65

ctors

-0.64

 omission

-0.64

JECT

-0.64

POSITIVE LOGITS

 flared

1.13

 tensions

1.05

 simmer

0.94

 flare

0.92

 tension

0.91

 escal

0.89

 cooker

0.88

 escalate

0.86

 raging

0.86

 arising

0.85

Activations Density 0.061%

phrases related to escalating conflicts and tensions between countries or groups

No Comments

No Known Activations