INDEX

Explanations

phrases indicating comparisons and competitive scenarios

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

 Pacific

-0.72

 Miko

-0.71

uclear

-0.62

 Koch

-0.59

 Ducks

-0.59

Kot

-0.58

 Nanto

-0.58

 Gong

-0.58

 Warm

-0.57

Tik

-0.57

POSITIVE LOGITS

*.

0.76

©¶æ¥µ

0.76

 attest

0.72

 emulate

0.70

 aspire

0.67

 deem

0.67

 accustomed

0.63

 theirs

0.62

igree

0.62

tical

0.62

Activations Density 0.213%

phrases indicating comparisons and competitive scenarios

No Comments

No Known Activations