INDEX

Explanations

names of people and specific details about their actions in different contexts

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs6144-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_6144

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

6,144

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

Shares

-0.67

idth

-0.63

atur

-0.63

 directions

-0.57

ruction

-0.56

":["

-0.54

 apologies

-0.54

how

-0.54

 Hanson

-0.53

 Balloon

-0.53

POSITIVE LOGITS

 abundantly

0.97

 easier

0.96

 happen

0.84

 seem

0.78

 harder

0.77

 impossible

0.76

 safer

0.75

unes

0.74

 tougher

0.73

 worthwhile

0.73

Activations Density 0.142%

names of people and specific details about their actions in different contexts

No Comments

No Known Activations