INDEX

Explanations

terms related to combat or physical abilities

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

ãĥĪ

-0.85

ãĥīãĥ©ãĤ´ãĥ³

-0.82

ãĥīãĥ©

-0.77

PsyNetMessage

-0.74

ãĥĩ

-0.72

ãĥŃ

-0.70

ãĥĵ

-0.69

ãĥŁ

-0.68

ãĥ¡

-0.67

======

-0.65

POSITIVE LOGITS

lers

1.29

ety

1.23

leground

1.09

ett

1.08

erer

1.07

eting

1.05

ering

1.05

erers

1.04

eters

1.03

ler

1.03

Activations Density 3.013%

terms related to combat or physical abilities

No Comments

No Known Activations