INDEX

Explanations

references to formal agreements or decisions, termed as "resolutions."

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

rium

-0.75

heon

-0.73

spring

-0.73

rower

-0.70

endon

-0.70

inen

-0.68

elong

-0.67

RN

-0.67

FactoryReloaded

-0.66

anooga

-0.66

POSITIVE LOGITS

 resolutions

0.88

 authorizing

0.87

 resolution

0.86

 Resolution

0.85

witz

0.78

 resolves

0.77

olution

0.73

ignty

0.72

mary

0.72

ancing

0.71

Activations Density 0.024%

references to formal agreements or decisions, termed as "resolutions."

No Comments

No Known Activations