INDEX

Explanations

phrases that emphasize the concept of news reporting or significant events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

awa

-0.77

KK

-0.72

respective

-0.71

wagon

-0.69

PB

-0.67

 supervised

-0.67

onne

-0.67

 Canaver

-0.67

netflix

-0.65

aceous

-0.64

POSITIVE LOGITS

 Errors

0.89

 Hours

0.74

 Herald

0.74

 Seasons

0.72

 Liberties

0.67

Times

0.66

requency

0.65

 Inquiry

0.64

 Planet

0.63

 Ages

0.63

Activations Density 0.020%

phrases that emphasize the concept of news reporting or significant events

No Comments

No Known Activations