INDEX

Explanations

repeated patterns or sequences of words within longer texts

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

 sacrific

-0.78

atography

-0.73

ources

-0.69

otor

-0.68

 dedicate

-0.67

peed

-0.67

aband

-0.65

poon

-0.64

 Leilan

-0.64

oided

-0.64

POSITIVE LOGITS

 pattern

0.84

 fallacy

0.83

pattern

0.81

 occurs

0.80

 manifested

0.80

 phenomenon

0.80

 happens

0.77

 repeats

0.77

 repeated

0.76

 occurring

0.76

Activations Density 0.338%

repeated patterns or sequences of words within longer texts

No Comments

No Known Activations