INDEX

Explanations

self-referential words indicating action or change happening

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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

SPONSORED

-0.76

ITNESS

-0.65

elected

-0.65

interstitial

-0.63

ifted

-0.62

aug

-0.60

 Married

-0.60

assadors

-0.60

ilingual

-0.59

onnaissance

-0.59

POSITIVE LOGITS

 balance

0.85

 fortunes

0.85

 unnecessarily

0.84

 altogether

0.84

 prematurely

0.84

 momentum

0.83

 inhib

0.83

 entire

0.82

 reins

0.82

 boundaries

0.81

Activations Density 7.666%

self-referential words indicating action or change happening

No Comments

No Known Activations