INDEX

Explanations

references to discussions or negotiations between parties

oai_token-act-pair · gpt-4o-mini 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ribution

-0.72

inion

-0.71

olls

-0.68

alach

-0.68

osate

-0.65

Reader

-0.65

ãĥ©ãĥ³

-0.64

Rah

-0.64

Notable

-0.62

Sensor

-0.61

POSITIVE LOGITS

 involving

0.95

 underway

0.93

pace

0.91

 commenced

0.90

 negotiations

0.90

hops

0.88

 between

0.84

hips

0.82

paces

0.82

Phase

0.81

Activations Density 0.047%

references to discussions or negotiations between parties

No Comments

No Known Activations