INDEX

Explanations

roles and responsibilities in various contexts, such as leadership roles, specific duties, and physical spaces used for different purposes

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

Exit

-0.77

ocumented

-0.70

tu

-0.70

isible

-0.69

uit

-0.67

ivers

-0.66

rehensive

-0.66

aware

-0.66

gru

-0.66

encers

-0.65

POSITIVE LOGITS

 mantle

1.34

 reins

1.05

 role

1.05

 roles

0.97

 responsibilities

0.96

 slack

0.94

 guise

0.93

 duties

0.90

 vacated

0.90

 responsibility

0.89

Activations Density 0.161%

roles and responsibilities in various contexts, such as leadership roles, specific duties, and physical spaces used for different purposes

No Comments

No Known Activations