INDEX

Explanations

statements or quotations made by people

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

estern

-0.95

ammy

-0.71

avorite

-0.71

peg

-0.70

OUP

-0.67

transfer

-0.66

esc

-0.66

ãĤ¼ãĤ¦ãĤ¹

-0.65

ynam

-0.65

asonic

-0.65

POSITIVE LOGITS

"[

0.87

 they

0.85

it

0.84

 "...

0.77

"(

0.75

"'

0.74

 "âĢ¦

0.73

 goodbye

0.72

 instead

0.70

 otherwise

0.70

Activations Density 0.078%

statements or quotations made by people

No Comments

No Known Activations