INDEX

Explanations

references to specific individuals, notes, and interactions in a conversational context

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

eret

-0.62

oriented

-0.58

ussen

-0.58

visory

-0.58

paralle

-0.57

inational

-0.57

egu

-0.57

ifies

-0.56

entimes

-0.56

 strengthens

-0.53

POSITIVE LOGITS

 glorious

0.76

 smug

0.74

 Cthulhu

0.73

 bliss

0.71

 sanity

0.71

 wretched

0.71

 mortals

0.71

 goddamn

0.70

 souls

0.69

 Patreon

0.67

Activations Density 1.631%

references to specific individuals, notes, and interactions in a conversational context

No Comments

No Known Activations