INDEX

Explanations

narrative descriptions involving personal experiences

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

quartered

-0.65

 overseen

-0.65

anwhile

-0.64

 oversaw

-0.56

 bolstered

-0.56

 broadly

-0.56

 facilitated

-0.56

 outset

-0.55

 controvers

-0.55

 nascent

-0.54

POSITIVE LOGITS

 fuckin

1.17

;)

1.12

 fucking

1.12

 haha

1.11

 kinda

1.10

 shit

1.10

 gonna

1.06

lol

1.06

 crap

1.05

:)

1.04

Activations Density 11.245%

narrative descriptions involving personal experiences

No Comments

No Known Activations