INDEX

Explanations

number sequences

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

 Bots

-0.81

 Pengu

-0.78

ela

-0.78

mosp

-0.78

mb

-0.74

 Âł Âł Âł Âł

-0.74

 Pascal

-0.73

BET

-0.73

brance

-0.72

ibo

-0.72

POSITIVE LOGITS

1.46

1.18

1.17

1.12

XII

1.06

0.96

0.95

0.94

ł

0.91

 twelve

0.88

Activations Density 0.071%

number sequences

No Comments

No Known Activations