INDEX

Explanations

words related to movement or change in status, specifically centered around 'returning' or 'reapplying'

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

uchs

-0.63

ahi

-0.61

iop

-0.60

anooga

-0.60

holes

-0.59

mouth

-0.59

hole

-0.57

PF

-0.56

 disse

-0.56

nut

-0.56

POSITIVE LOGITS

ITED

0.95

din

0.84

 recapt

0.81

prise

0.79

 refreshed

0.74

issance

0.74

ees

0.72

osponsors

0.72

owship

0.71

issions

0.71

Activations Density 1.730%

words related to movement or change in status, specifically centered around 'returning' or 'reapplying'

No Comments

No Known Activations