INDEX

Explanations

cities or geographic locations

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

wcs

-0.77

CEO

-0.69

*/(

-0.69

APD

-0.67

 relate

-0.64

 reperto

-0.64

 subscript

-0.63

hole

-0.61

 repertoire

-0.60

 subscrib

-0.60

POSITIVE LOGITS

Sao

0.90

 Gren

0.88

 Janeiro

0.83

Ing

0.83

Tir

0.83

 Calais

0.83

 Alexandria

0.83

Nev

0.82

 Trinidad

0.81

 Yugoslavia

0.80

Activations Density 0.694%

cities or geographic locations

No Comments

No Known Activations