INDEX

Explanations

references to innocence or lack of involvement in various situations

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

duction

-0.67

hooting

-0.65

idding

-0.64

Os

-0.64

FIG

-0.63

 Digest

-0.62

early

-0.62

GE

-0.62

uberty

-0.61

 Supplement

-0.61

POSITIVE LOGITS

 knack

0.92

 doubts

0.88

 vested

0.88

 recourse

0.85

 reservations

0.84

 advantages

0.84

 regrets

0.83

 clout

0.83

 sympathy

0.83

 chance

0.83

Activations Density 0.214%

references to innocence or lack of involvement in various situations

No Comments

No Known Activations