INDEX

Explanations

mentions of specific names or terms related to particular topics

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

eers

-0.78

SHIP

-0.69

 Gent

-0.68

 Wand

-0.68

eer

-0.66

front

-0.66

 Clover

-0.65

Ĥ¬

-0.65

kered

-0.63

 trumpet

-0.62

POSITIVE LOGITS

acker

1.13

acking

1.11

udence

1.10

ussia

1.09

ut

1.06

abbit

1.05

aternity

1.05

anca

1.04

inkle

1.04

acket

1.03

Activations Density 0.701%

mentions of specific names or terms related to particular topics

No Comments

No Known Activations