INDEX

Explanations

instances where the writer emphasizes a high quantity or intensity of something

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

ħĭ

-0.81

saf

-0.80

pherd

-0.76

selves

-0.75

othy

-0.75

Ń·

-0.74

²¾

-0.73

swer

-0.71

byn

-0.70

İĭ

-0.70

POSITIVE LOGITS

 emotion

0.81

 firepower

0.81

 hype

0.79

 crap

0.75

 shit

0.75

dmg

0.74

 attention

0.74

fun

0.73

 negativity

0.72

 stuff

0.71

Activations Density 0.031%

instances where the writer emphasizes a high quantity or intensity of something

No Comments

No Known Activations