INDEX

Explanations

information related to political events or discussions

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

ceptions

-0.69

 Surve

-0.64

ernels

-0.63

unts

-0.63

 lines

-0.62

 glare

-0.62

lees

-0.60

chell

-0.60

luster

-0.59

idents

-0.59

POSITIVE LOGITS

 titled

0.93

 whereby

0.91

 wherein

0.87

 which

0.82

 consisting

0.75

icial

0.74

which

0.71

 sorts

0.69

ographed

0.69

 Called

0.68

Activations Density 3.249%

information related to political events or discussions

No Comments

No Known Activations