INDEX

Explanations

information related to events or actions that have caused outrage or concern in a community

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

tyr

-0.79

 Brach

-0.73

 millenn

-0.61

 Mock

-0.60

-0.58

 Inqu

-0.58

 arsen

-0.57

avorite

-0.56

Mub

-0.56

 etched

-0.55

POSITIVE LOGITS

stretched

1.33

fitted

1.21

casts

1.14

dated

1.14

smart

1.13

doors

1.12

skirts

1.12

lier

1.09

fitting

1.07

lying

1.07

Activations Density 0.048%

information related to events or actions that have caused outrage or concern in a community

No Comments

No Known Activations