INDEX

Explanations

short timeframes and events that occur within them

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

acus

-0.74

Reviewer

-0.68

ahime

-0.64

activity

-0.62

orescence

-0.61

usage

-0.60

olas

-0.60

eret

-0.59

abwe

-0.59

inite

-0.59

POSITIVE LOGITS

ago

1.13

shy

0.98

 ahead

0.96

Ago

0.89

 kilometers

0.88

 away

0.87

 apart

0.85

 kilometres

0.83

 distance

0.82

 miles

0.82

Activations Density 3.057%

short timeframes and events that occur within them

No Comments

No Known Activations