INDEX

Explanations

phrases related to observation and discovery

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_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

umar

-0.79

ella

-0.73

ovie

-0.72

ature

-0.69

uli

-0.64

icus

-0.64

inqu

-0.64

PHOTOS

-0.62

cius

-0.62

Phill

-0.61

POSITIVE LOGITS

 plenty

1.00

 fewer

0.91

 lots

0.90

 tons

0.86

 yourselves

0.85

 yourself

0.81

 references

0.79

 alot

0.75

 MUCH

0.74

 ample

0.70

Activations Density 0.119%

phrases related to observation and discovery

No Comments

No Known Activations