INDEX

Explanations

brand-related content

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

abama

-0.81

veyard

-0.77

ulhu

-0.76

cale

-0.73

DAY

-0.69

isites

-0.68

pmwiki

-0.68

isite

-0.65

nesota

-0.65

tein

-0.65

POSITIVE LOGITS

ishing

1.00

enburg

1.00

 loyalty

0.97

Brand

0.87

ishes

0.81

ished

0.80

brand

0.79

ages

0.79

eli

0.75

 logos

0.75

Activations Density 0.759%

brand-related content

No Comments

No Known Activations