INDEX

Explanations

elements and instances of self-reflection and existential questioning

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

 brid

-0.81

 increasingly

-0.78

 raft

-0.78

 pressing

-0.78

 revers

-0.77

 favour

-0.76

 continuous

-0.75

 favor

-0.75

 coral

-0.74

 cycl

-0.73

POSITIVE LOGITS

And

1.51

Advertisements

1.42

It

1.40

They

1.40

Instead

1.38

Because

1.38

That

1.37

However

1.36

Anyone

1.36

Until

1.35

Activations Density 0.401%

elements and instances of self-reflection and existential questioning

No Comments

No Known Activations