INDEX

Explanations

references to geographic locations, particularly country names and markets

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

icer

-0.75

ashington

-0.74

acl

-0.73

zik

-0.69

atism

-0.67

apse

-0.67

alin

-0.66

struct

-0.66

cler

-0.66

msg

-0.66

POSITIVE LOGITS

 shores

0.97

 Territories

0.91

 markets

0.89

 indie

0.85

NTS

0.83

 localization

0.81

 territories

0.80

 distributors

0.79

 distributor

0.79

 Telecom

0.79

Activations Density 0.178%

references to geographic locations, particularly country names and markets

No Comments

No Known Activations