INDEX

Explanations

phrases related to locations, events, and businesses

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gradient

-0.69

acon

-0.68

Query

-0.67

uated

-0.65

isson

-0.64

BIL

-0.64

Tweet

-0.62

Ã¨re

-0.62

 wondered

-0.61

_-

-0.60

POSITIVE LOGITS

 world

1.25

 vicinity

1.09

 shortest

1.08

 country

1.06

 entire

1.06

 midst

1.05

 nation

1.04

 universe

0.99

 hemisphere

0.95

 history

0.95

Activations Density 0.107%

phrases related to locations, events, and businesses

No Comments

No Known Activations