INDEX

Explanations

phrases related to a majority sentiment or decision

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uers

-0.66

thur

-0.65

LV

-0.63

KC

-0.63

rir

-0.61

æ©

-0.60

 Gutenberg

-0.60

estamp

-0.59

ny

-0.58

åĤ

-0.58

POSITIVE LOGITS

(>

0.86

majority

0.85

 opinion

0.85

 shareholder

0.77

OTUS

0.76

 majority

0.75

ority

0.71

(~

0.71

 majorities

0.69

orescence

0.69

Activations Density 0.659%

phrases related to a majority sentiment or decision

No Comments

No Known Activations