INDEX

Explanations

future predictive statements

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.66

JD

-0.62

engineering

-0.62

amps

-0.61

HQ

-0.61

quickShipAvailable

-0.59

 Modified

-0.58

hea

-0.58

 Sandwich

-0.57

 Chal

-0.56

POSITIVE LOGITS

be

1.36

 continue

1.21

 gladly

1.20

 undoubtedly

1.11

 doubtless

1.09

 eventually

1.07

 inevitably

1.06

 surely

1.05

 arrive

1.03

 soon

1.02

Activations Density 1.185%

future predictive statements

No Comments

No Known Activations