INDEX

Explanations

specific patterns related to descriptions or narratives involving individuals interacting and communicating

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

itiveness

-0.63

 issu

-0.62

 DISTRICT

-0.57

 Releases

-0.55

artments

-0.55

itizens

-0.54

REG

-0.54

 Explos

-0.53

ISO

-0.53

 RELE

-0.53

POSITIVE LOGITS

 whom

1.10

his

0.91

him

0.88

 hers

0.85

 himself

0.82

he

0.79

his

0.79

him

0.77

 teammate

0.76

 classmate

0.72

Activations Density 0.880%

specific patterns related to descriptions or narratives involving individuals interacting and communicating

No Comments

No Known Activations