INDEX

Explanations

terms related to social interactions, alliances, and relationships

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

authorized

-0.73

overe

-0.71

ensibly

-0.67

grim

-0.67

printed

-0.66

industrial

-0.66

 Cout

-0.66

acco

-0.65

omsday

-0.65

atism

-0.64

POSITIVE LOGITS

lies

1.03

liness

0.97

hips

0.92

lier

0.88

liest

0.86

ships

0.80

ship

0.78

 colleagues

0.77

 colleague

0.76

hesis

0.74

Activations Density 1.935%

terms related to social interactions, alliances, and relationships

No Comments

No Known Activations