INDEX

Explanations

situations where clarity or explicitness is emphasized

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

 Niet

-0.65

 horizont

-0.63

rought

-0.59

cheon

-0.57

idth

-0.57

 exceeded

-0.57

rity

-0.56

verages

-0.56

Root

-0.56

 spearheaded

-0.55

POSITIVE LOGITS

 disappear

0.80

 noises

0.79

 impression

0.77

 happen

0.76

 debut

0.76

 mistake

0.76

sense

0.75

ends

0.73

 Murd

0.72

 noise

0.71

Activations Density 1.243%

situations where clarity or explicitness is emphasized

No Comments

No Known Activations