INDEX

Explanations

significant years, particularly those related to events and changes in various contexts

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

ERY

-0.62

ACTION

-0.61

needed

-0.61

rored

-0.61

IONS

-0.61

GoldMagikarp

-0.60

venge

-0.60

aliation

-0.56

BLIC

-0.55

require

-0.54

POSITIVE LOGITS

 onwards

1.84

 onward

1.61

 inception

0.98

 downwards

0.93

 outset

0.90

 standpoint

0.87

 perspective

0.85

ratch

0.76

 viewpoint

0.76

 till

0.76

Activations Density 0.058%

significant years, particularly those related to events and changes in various contexts

No Comments

No Known Activations