INDEX

Explanations

references to specific individuals or characters, potentially related to news or entertainment

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_fs12288-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

12,288

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

cles

-0.95

BLE

-0.80

lesiastical

-0.79

nesia

-0.77

opathic

-0.76

ntil

-0.72

opathy

-0.72

fortune

-0.71

ilar

-0.70

ggle

-0.70

POSITIVE LOGITS

 Stewart

0.82

 Olsen

0.71

ards

0.70

 Rhodes

0.69

rey

0.68

yard

0.67

 Chamberlain

0.67

 Gardner

0.65

 Downing

0.65

rade

0.64

Activations Density 0.177%

references to specific individuals or characters, potentially related to news or entertainment

No Comments

No Known Activations

references to specific individuals or characters, potentially related to news or entertainment

No Comments

No Known Activations