INDEX

Explanations

mentions of accomplishments or achievements in various 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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

 Mechdragon

-0.82

hement

-0.75

 Guang

-0.70

 Slug

-0.68

XII

-0.66

 Calculator

-0.65

 Shen

-0.65

 Krypt

-0.64

 Lash

-0.64

Ide

-0.63

POSITIVE LOGITS

shall

0.99

cause

0.89

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

0.89

ever

0.88

enough

0.87

together

0.86

due

0.86

ÂłÂłÂłÂłÂłÂłÂłÂł

0.85

accompan

0.84

ó

0.84

Activations Density 0.725%

mentions of accomplishments or achievements in various fields

No Comments

No Known Activations