INDEX

Explanations

words related to forms of medical conditions or treatments

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Hots

-0.66

 VIDEOS

-0.64

selves

-0.57

 spot

-0.57

ghan

-0.56

itness

-0.56

 Bars

-0.56

itbart

-0.56

 grill

-0.56

ĸļ

-0.56

POSITIVE LOGITS

aldehyde

1.33

ative

1.10

idable

1.08

ality

1.04

atter

1.00

atted

0.98

ulating

0.95

ul

0.94

fitting

0.93

ulator

0.93

Activations Density 0.020%

words related to forms of medical conditions or treatments

No Comments

No Known Activations