INDEX

Explanations

detailed descriptions of events and situations

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

olves

-0.78

SourceFile

-0.76

 Coverage

-0.66

zing

-0.64

externalActionCode

-0.63

nexus

-0.62

aukee

-0.61

clave

-0.61

ORPG

-0.60

ÅĤ

-0.60

POSITIVE LOGITS

 alas

1.21

 moreover

1.00

 secondly

0.98

 interestingly

0.97

 unsurprisingly

0.96

cru

0.96

uh

0.93

 furthermore

0.92

yes

0.91

 frankly

0.91

Activations Density 0.479%

detailed descriptions of events and situations

No Comments

No Known Activations