INDEX

Explanations

phrases related to recognizing achievements, accolades, and notable contributions in different fields

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

 Refresh

-0.69

ouri

-0.66

 Grey

-0.63

 Dund

-0.61

 Univ

-0.61

oresc

-0.60

ct

-0.60

cel

-0.58

 exception

-0.57

 Torch

-0.57

POSITIVE LOGITS

 also

0.88

 actively

0.86

also

0.85

 ALSO

0.81

 actually

0.77

 secondly

0.74

actually

0.73

yon

0.72

 downright

0.72

sic

0.71

Activations Density 2.149%

phrases related to recognizing achievements, accolades, and notable contributions in different fields

No Comments

No Known Activations