INDEX

Explanations

references to the term "world" in various contexts

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

ippi

-0.90

edin

-0.82

iple

-0.78

veyard

-0.77

utic

-0.76

»Ĵ

-0.75

ifer

-0.71

ibility

-0.69

ilipp

-0.68

dq

-0.68

POSITIVE LOGITS

wide

1.13

building

0.90

 renown

0.87

 championships

0.85

builders

0.85

 religions

0.84

Wide

0.83

 domination

0.82

 renowned

0.81

views

0.79

Activations Density 0.032%

references to the term "world" in various contexts

No Comments

No Known Activations