INDEX

Explanations

phrases related to the concept of "initial impressions."

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

rolet

-0.76

 Journals

-0.71

rored

-0.67

 Calories

-0.67

ux

-0.65

vance

-0.64

 Gould

-0.61

ufact

-0.61

 à¤

-0.61

 Rolls

-0.59

POSITIVE LOGITS

 glance

1.57

 blush

1.55

 sight

1.04

 hesitant

0.91

 responders

0.81

 reluctant

0.79

hes

0.76

 confused

0.74

 glimpse

0.73

 puzz

0.72

Activations Density 0.078%

phrases related to the concept of "initial impressions."

No Comments

No Known Activations