INDEX

Explanations

phrases related to police work and investigations

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

horn

-0.73

owitz

-0.73

wit

-0.71

burgh

-0.71

kers

-0.70

oras

-0.68

rican

-0.68

emi

-0.67

iva

-0.67

bern

-0.67

POSITIVE LOGITS

 routine

1.15

 maintenance

1.11

 routines

1.00

ized

0.95

ised

0.91

 checks

0.84

ALLY

0.82

 tasks

0.82

cy

0.81

 regimen

0.81

Activations Density 0.027%

phrases related to police work and investigations

No Comments

No Known Activations