INDEX

Explanations

instances of acronyms and abbreviations related to information systems

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

ruciating

-0.87

ãĥĥãĥĪ

-0.76

boy

-0.69

ohydrate

-0.67

ilight

-0.67

isting

-0.67

 Pradesh

-0.65

ised

-0.64

elta

-0.64

 msec

-0.64

POSITIVE LOGITS

TRY

1.22

KI

1.00

BN

0.99

SI

0.98

NER

0.95

CO

0.95

KO

0.94

TS

0.94

PLAY

0.94

COL

0.94

Activations Density 0.008%

instances of acronyms and abbreviations related to information systems

No Comments

No Known Activations