INDEX

Explanations

verbs related to driving and motivation

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Seym

-0.96

ereo

-0.76

ertain

-0.72

yip

-0.72

ellen

-0.69

umbn

-0.69

Lum

-0.69

anamo

-0.68

ileaks

-0.68

iannopoulos

-0.66

POSITIVE LOGITS

train

0.92

driving

0.90

 driving

0.87

wheel

0.82

 dealership

0.74

 driven

0.74

 Driving

0.73

club

0.71

 wedge

0.70

driver

0.70

Activations Density 0.550%

verbs related to driving and motivation

No Comments

No Known Activations