INDEX

Explanations

concepts related to conflict and power dynamics

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

 referen

-0.82

Liv

-0.70

 Coat

-0.63

 Socket

-0.63

 trademark

-0.63

ONSORED

-0.62

ITNESS

-0.59

RIPT

-0.59

 Veterinary

-0.59

ammy

-0.58

POSITIVE LOGITS

lessness

1.02

fulness

1.02

liness

0.94

lessly

0.89

flies

0.87

ankind

0.87

utics

0.84

making

0.82

istically

0.82

fully

0.81

Activations Density 0.277%

concepts related to conflict and power dynamics

No Comments

No Known Activations