INDEX

Explanations

words related to television shows and their characters

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

 Vill

-0.67

Ô

-0.65

 Assignment

-0.64

imon

-0.63

 Verse

-0.63

 Dhabi

-0.62

NetMessage

-0.61

 Teams

-0.60

sbm

-0.59

annie

-0.59

POSITIVE LOGITS

 itself

1.12

's

0.95

osphere

0.91

iverse

0.85

wide

0.83

runner

0.81

runners

0.79

ultimate

0.77

 hierarchy

0.76

wright

0.72

Activations Density 1.542%

words related to television shows and their characters

No Comments

No Known Activations