INDEX

Explanations

words related to entertainment media such as TV shows, movies, actors, and genres

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

-0.79

orio

-0.72

 Aden

-0.71

lease

-0.71

çĦ

-0.70

andum

-0.70

 Cabrera

-0.68

Aad

-0.66

 sacked

-0.66

 Aadhaar

-0.66

POSITIVE LOGITS

 enthusiast

1.16

 themed

1.16

 enthusiasts

1.12

 thriller

1.11

 oriented

1.07

 classics

1.07

 realism

1.03

 genre

1.02

 genres

1.02

 sensibilities

1.01

Activations Density 1.725%

words related to entertainment media such as TV shows, movies, actors, and genres

No Comments

No Known Activations