INDEX

Explanations

verbal actions related to influencing or controlling 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_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

 preferring

-0.72

 Invasion

-0.61

 Cruiser

-0.60

 wander

-0.58

 Originally

-0.58

=>

-0.57

 Birthday

-0.57

 nervously

-0.56

 inquired

-0.56

Flo

-0.55

POSITIVE LOGITS

balance

0.86

rid

0.81

uate

0.78

tle

0.73

down

0.73

ãĥķãĤ¡

0.70

 forth

0.70

itial

0.70

ify

0.69

igate

0.68

Activations Density 0.265%

verbal actions related to influencing or controlling situations

No Comments

No Known Activations