INDEX

Explanations

information related to time references

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

 representation

-0.64

resa

-0.60

 knots

-0.59

Mog

-0.58

oat

-0.58

 Musk

-0.58

 Monstrous

-0.58

emis

-0.56

ged

-0.56

 Stall

-0.55

POSITIVE LOGITS

adays

1.09

here

0.87

 defunct

0.80

abouts

0.79

ogene

0.73

tainment

0.70

forth

0.70

orea

0.69

heric

0.68

èª

0.67

Activations Density 0.026%

information related to time references

No Comments

No Known Activations